ÿWPCP ûÿ2BVJ Z ÿÿ úCourierTÿÿ‰?xxx°p£Ûxþ6X@ÉüKX@þþþþþþþÿþÿÿÿþÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿHP LaserJet 4 PostScriptHPLA4POS.PRSÛx Œ @ɇÏhhhhð' CX@ÐÐûÿ2ÿÿ6>N|x3|xÐÐ Because of technical limitations, charts and footnotes were dropped from this document. Before the FEDERAL COMMUNICATIONS COMMISSION FCC 94-38 Washington, D.C. 20554 In the Matter of ) ) Implementation of Sections of ) the Cable Television Consumer ) Protection and Competition ) MM Docket No. 92-266 Act of 1992: Rate Regulation ) SECOND ORDER ON RECONSIDERATION, FOURTH REPORT AND ORDER, AND FIFTH NOTICE OF PROPOSED RULEMAKING Adopted: February 22, 1994 ; Released: March 30, 1994 By the Commission: Commissioner Barrett issuing a statement. Comment Date: June 29, 1994 Reply Comment Date: July 29, 1994 Table of Contents Paragraphs: I. Introduction 1 II. Second Order on Reconsideration 9 A. Executive Summary 9 B. Regulations Governing Rates of Basic and Cable Programming Service Tiers 43 1. Background 43 2. Impact on the National Economy 54 3. Estimating the Competitive Differential 67 a. Disaggregated Treatment of Low Penetration, Overbuild, and Municipal Systems 72 b. Improved Variables 77 c. Tests for the Effects of CostÔ"h)0*0*0*°°°1ð"ÔŒ and Demand Differences 81 d. System Size and the Competitive Differential 86 e. The New Competitive Differential 90 4. Applying the Competitive Differential 106 a. Systems Not Entitled to Transition Relief 114 b. Systems Entitled to Transition Relief 117 (i) Systems Owned by Small Operators 117 (ii) Low-Priced Systems 123 c. Application of Transition Relief 127 d. Regulated Rates at the End of the Transition Period 132 e. Calculation of Refund Liability 135 f. Notice to Subscribers 139 g. Procedural Issues for Franchising Authorities 144 h. Pending Complaints Before the Commission 150 5. Commission Authority to Adopt the Modified Ratemaking Approach 153 6. The Price Cap Governing Cable Service Rates 169 a. Calculation of External Costs 169 b. Copyright Fees 178 c. Pole Attachment Fees 181 7. Other Rate Issues 183 a. Commercial Rates 184 b. Rate Relief for Alaska and Hawaii 186 c. Basic Tier Access Charge 189 C. "A La Carte" Packages 191 D. Small Systems 201 1. Small System Administrative Relief 201 a. Background 201 (i) Streamlined Rate Reductions 208 (ii) Company-Wide Averaging of Equipment 218 (iii)Other Proposals for Administrative 222 Relief 2. Headend vs. Franchise Area Definition of Small Systems 226 3. Termination of Rate Regulation Stay for Small Systems 228 III. Fourth Report and Order 229 A. Introduction 229 B. Adjustments to Capped Rates for Addition and Deletion of Channels 231 1. Background 231 2. Regulatory Goals 236 3. The Going-Forward Methodology 239 C. Upgrades Initiated Shortly Before Rate Regulation 250 IV. Fifth Notice of Proposed Rulemaking 254 V. Regulatory Flexibility Act Analysis 258 A. Final Analysis for the Fourth Report and Order and Second Order on Reconsideration 258 B. Initial Regulatory Flexibility Analysis for the Fifth Notice of Proposed Rulemaking 263 VI. Paperwork Reduction Act 271 VII. Procedural Provisions 272 VIII. Ordering Clauses 274 Appendices I. Introduction 1. In the April 1993 Report and Order and Further Notice of Proposed Rule Making ("the Rate Order") in this docket, the Commission adopted cable rate regulation rules and policies pursuant to the Cable Television Consumer Protection and Competition Act of 1992. In the Rate Order, we adopted a "benchmark" approach for setting initial rates for regulated cable service. Under the benchmark approach, regulated cable systems were required to use a formula established in the Rate Order to calculate an applicable benchmark -- an estimate of the rate that a cable system subject to effective competition with similar characteristics would charge. Rates of cable systems at or below the benchmark were presumed to be reasonable; rates above the benchmark were presumed to be unreasonable. Cable systems whose rates exceeded the applicable benchmark were required to reduce their rates either to the benchmark or by ten percent, whichever reduction was less. This ten percent "competitive differential" represented the average difference that the Commission determined existed between the rates of competitive and non-competitive cable systems. 2. In this decision, we amend our cable rate regulations to ensure both that the rates consumers pay for regulated cable services are reasonable and that our rules continue to promote economic growth in the cable industry. Using a revised economic model, we have recalculated the competitive differential and have concluded that the 17 percent differential determined by the revised model more accurately estimates the difference between effectively competitive and noncompetitive cable rates than the ten percent differential established in the Rate Order. 3. In addition, we have reconsidered our benchmark approach. As the record stands, we believe that all regulated cable systems should be required to establish rates based on the revised competitive differential unless they justify other rates through a cost-of-service showing. We conclude that this result is warranted for several reasons, including the language of the statute, economic theory, and the legislative history of the 1992 Cable Act establishing that cable systems generally have had an opportunity to exercise market power. Furthermore, the Commission's data continues to demonstrate that rates charged by systems that are not subject to effective competition, as defined in the statute, are generally higher than rates charged by systems facing effective competition. We believe, therefore, that our revised competitive differential best estimates the extent to which noncompetitive systems have been charging unreasonable rates. 4. We are, however, establishing special transition rules that provide that noncompetitive cable systems charging relatively low prices (as measured by a revised benchmark that incorporates the 17 percent competitive differential) will not have to reduce their rates by the full competitive differential until the Commission has collected and analyzed data about such operators' prices and costs, and determined whether such a reduction is inappropriate. We anticipate that some data relevant to this issue will be collected as part of an industry cost study to be conducted by the Commission, while other data will be submitted by affected cable systems. At the conclusion of this process, the regulated rates of such systems will be set to reflect the full 17 percent differential if our analysis does not show that the resulting rates would be unreasonably low -- that is, the rates would be lower than they would be if set by competitive pressures as determined by cost comparisons between noncompetitive systems and systems subject to effective competition. 5. We are following a similar approach with respect to cable systems owned by small cable operators, defined as cable companies that serve a total subscriber base of 15,000 or fewer. We will not require small operators to apply the new competitive differential until the Commission has collected and analyzed data about such operators' prices and costs. As with systems that charge relatively low prices, we anticipate that some data relevant to this issue will be collected as part of an industry cost study to be conducted by the Commission, while other data will be submitted by affected cable systems. At the conclusion of our analysis, small operators will have to apply the full competitive differential in the absence of evidence showing that its application would set their rates at unreasonable levels based on any unique costs that are systematically experienced by small operators, as well as differences demonstrated between noncompetitive and competitive small operators. 6. In addition, we adopt a Fourth Report and Order establishing a "going-forward" mechanism to govern future rate adjustments resulting from channel additions or deletions, or system upgrades. We also issue a Fifth Notice of Proposed Rulemaking to seek comment on whether and how our benchmark methodology should apply to systems with more than 100 channels. In addition, we seek comment on whether the going-forward rules adopted in this Order should be modified for systems with more than 100 channels and, more generally, whether our going-forward methodology should be modified to provide greater or lesser compensation to operators when channels are added or deleted from regulated tiers. We also seek comment on whether we should allow cable operators to charge higher rates for cable service provided to commercial establishments than to residential households. Finally, we seek comment on whether we should adopt further requirements to govern systems' move from transition relief to full reduction rates. 7. Simultaneously, we are issuing two related orders. One order resolves a variety of cable rate regulation issues that do not relate to rate calculations. The other order establishes interim rules governing the "cost-of-service" procedures that cable operators may invoke if they believe that their costs of providing regulated cable service will support rates that are higher than those produced by the revised benchmark approach adopted in this Order. 8. As a result of these revised rules, most regulated cable operators will have to apply the revised competitive differential by May 15, 1994, or, pursuant to some conditions, by July 14, 1994. A limited number of operators will be eligible for special transition treatment pursuant to which they will make smaller or no, rate reductions pending completion of price/cost studies by the Commission. Cable systems that are eligible for transition treatment will be subject to the full 17 percent reduction later, unless our price/cost analysis reveals that its application to these systems is inappropriate. II. Second Order on Reconsideration A. Executive Summary 9. After Congress deregulated most cable rates in 1984, the rates for many subscribers rose substantially. Congress proceeded to study the cable industry, and found that "most cable television subscribers have no opportunity to select between competing cable systems." The result, it determined, "is undue market power for the cable operator as compared to that of consumers." This conclusion is supported by the findings of numerous economists. 10. In the 1992 Cable Act, Congress determined that the rates charged by cable systems that face head-to-head competition should not be regulated. Congress further decided that cable systems operated by municipalities and cable systems with low penetration should be deemed, as a matter of law, to be subject to "effective competition," and hence not subject to rate regulation. But Congress provided that the rates charged by all other cable systems for basic tier service and upper program (i.e., cable service) tiers should be subject to regulation. 11. The 1992 Cable Act directs the Commission to adopt regulations to ensure that rates for the basic service tier are "reasonable." Congress made clear that the basic tier regulations must "be designed to achieve the goal of protecting subscribers" from paying rates that are higher than those that would be charged if their cable operator faced "effective competition." Congress also required the Commission to "take into account" seven different factors, most of which relate to the costs of providing, and the revenues derived from, basic service. While providing that basic tier rates must be "reasonable," establishing "the goal of protecting consumers," and directing the Commission to "take into account" seven different factors, Congress did not require the Commission to follow a specific methodology in setting rates. Indeed, the Conference Committee deleted a provision requiring a formulaic approach, and explained that it was giving "the Commission the authority to choose the best method of ensuring reasonable rates for the basic service tier." Congress followed a similar approach with respect to regulated upper tier service. 12. In April 1993, the Commission adopted cable rate regulations to implement these statutory provisions. At that time, we determined that the best approach for meeting our Congressional mandate was to rely principally on a benchmark methodology that requires noncompetitive systems to set rates at levels that approximate competitive prices. That conclusion was derived both from the language of the statute mandating reasonable rates and from economic theory, which states that prices charged in a competitive marketplace are reasonable for operators and subscribers alike. Competitive prices accurately reflect the true economic costs of the services provided. Such prices, thus, provide market signals to guide consumer purchasing decisions and producer investment choices at levels that lead to the greatest net benefit from the consumption of those services. 13. In April 1993, we further concluded that the benchmark methodology should be coupled with a cost-of-service process to permit operators to prove, on a case-by-case basis, that their costs support rates higher than those produced by the benchmark calculations. In addition, we adopted a "price cap" mechanism to ensure that future adjustments to initial rates set pursuant to the benchmark will be reasonable for both operators and subscribers. We also announced that we would (1) examine systems with rates substantially above competitive levels to determine whether their rates were justified by higher costs, and (2) seek to refine our analyses through further industry surveys. This comprehensive regulatory plan, we decided, addressed all of the various statutory factors and ensured that rates for all regulated service tiers will be reasonable, as required by the Act. 14. The petitions for reconsideration filed in this proceeding have given us an opportunity to undertake a comprehensive review of our rate regulation scheme. The evidence and analysis now before us support our earlier conclusion that the benchmark methodology, coupled with a cost-of-service process, is an appropriate means of setting initial regulated cable rates. At the same time, the record and further analysis also persuade us that we should modify our methodology so as best to set "reasonable" rates for cable systems that are not subject to effective competition, as defined in the statute. Accordingly, we adopt several important modifications to our current rate regulation approach. 15. In implementing the congressional directives on reconsideration, we are guided principally by economic theory, which states that a cable system that is not subject to competition generally will have an opportunity to charge a higher than competitive rate for its package of program services and equipment. As a result, in our view, a noncompetitive rate is not a "reasonable" rate. Rather, "reasonable" rates are lower than noncompetitive rates, and should reflect a price similar to that charged by a system that faces competition. 16. Moreover, upon further reflection, we find no conclusive basis in economic theory or the record to assume that a cable operator that is subject to little or no competition, and whose rates, nonetheless, appear to be low (i.e., at or below a benchmark), is not exercising market power. Market power is reflected in the gap between price and cost. Without more detailed information on cost and demand conditions, it is impossible to determine the amount by which an operator's revenues exceed its costs. An operator's rates may be relatively low on account of the cost or demand conditions it faces, rather than because the operator is not exercising market power. Our analysis indicates that behavior reflective of market power may exist generally within the noncompetitive sector of the cable industry. In this Order on Reconsideration, we calculate a new "competitive differential" -- our best estimate of the average amount by which the rates charged by a cable operator not facing effective competition exceeds "reasonable" rates -- and conclude that it should apply to all regulated cable systems. 17. However, because we do not have sufficient cost data at this time, we are establishing transition rules for some cable operators. First, operators with relatively low prices (as measured by their position vis-a-vis the revised benchmark) will not be required immediately to reduce rates by the new competitive differential. The full reduction for these operators will be deferred until the Commission has collected and analyzed data about such operators' prices and determined whether the full reduction is inappropriate. As indicated, some data on costs and prices will be collected as part of an industry cost study to be conducted by the Commission. Cable operators and other interested parties will be given the opportunity to present information on costs and prices to help provide a complete record. This will give cable operators the opportunity to submit evidence showing that cable systems charging rates at or below the new benchmark are not exercising market power and, therefore, might warrant an application of less than the full revised competitive differential. 18. Second, we will not require cable systems owned by small operators to apply the competitive differential immediately. We recognize that some small systems may face higher than average costs. We also recognize that small operators may not have the financial resources to sustain the impact of a significant rate reduction. Our cost study thus will also permit small operators to present evidence showing whether, or to what extent, our revised competitive differential should be applied to them due to (1) the presence of unique costs, as identified in the record, that are systematically experienced by small operators, and (2) any differences that may be demonstrated between noncompetitive and competitive small operators. We note, however, that the existing record does not enable us to find that small operators and relatively low priced operators should not eventually be required to apply the full competitive differential. They therefore will be required to apply the full differential unless our analysis of the price/cost data we collect persuades us that a smaller competitive differential should be applied to them. 19. As noted, we also adjust the competitive differential in this Order. We previously calculated the competitive differential as approximately ten percent. We made that calculation by analyzing data drawn from a sample of cable systems that included regulated systems and each of the three types of systems that Congress has exempted from rate regulation -- systems that have penetration rates less than 30 percent ("low penetration systems"), systems that face actual competition ("overbuilds"), and systems that are operated by municipalities ("municipals"). Our previous approach simply averaged the data from all systems subject to effective competition. On reconsideration, we have more closely analyzed the data from all three types of systems, and have used a qualitative, rather than arithmetic, analysis to determine the differential whose application best approximates the "reasonable" rate that would be charged by a system that faces effective competition. Our revised competitive differential of 17 percent also reflects our analysis of the various factors that Congress has instructed us to take into account. 20. Our refined approach will help ensure that consumers served by noncompetitive systems will be offered regulated services at reasonable rates. In addition, when the rates for regulated services fall as a result of the application of the competitive differential, the quantities of these services demanded by consumers should rise as new subscribers are added, fewer subscribers disconnect, and individual subscribers purchase additional programming services. These increases in demand will help offset the effect of the required rate reductions on operator revenues. The increased sales of existing regulated services should also be supplemented by demand for new services, both regulated and unregulated. 21. In particular, reducing regulated service rates will increase operators' motivation to invest in unregulated services that will apply advanced technology and to introduce entirely new services, such as broadband interactive services. Cable systems that now offer regulated service without competition will have an incentive to upgrade their systems with new capabilities and will have an incentive to introduce enhanced functions, such as interactivity, that are not subject to rate regulation. These developments will benefit operators by increasing revenues and benefit subscribers by increasing the affordability and availability of unregulated service offerings. 22. Our "going-forward" mechanism and cost-of-service procedures also will facilitate the development of new regulated services. For instance, because we are adopting a "going- forward" methodology for increasing rates when new program services are added to regulated service tiers, operators may launch new program services as part of regulated service tiers. A key concern expressed by operators and programmers throughout this proceeding has been that the benchmark approach may not permit operators to respond to marketplace incentives to expand the services included in regulated program tiers. The "going- forward" methodology set forth in this Order provides such incentives for the benefit of operators, programmers, and subscribers alike. 23. Our approach to rate-setting preserves incentives to invest in new services without imposing the burdens of financing new, unregulated offerings on regulated service subscribers. There is no sound policy reason to permit regulated operators to charge rates for regulated services that are higher than reasonable levels in order to support investment in plant used principally to provide unregulated services. Indeed, it has long been the Commission's policy to protect regulated ratepayers from subsidizing unregulated offerings through supra-competitive regulated rates. Rather, the risk of providing new, unregulated services has been placed, appropriately, on investors and shareholders. We see no reason to depart from that policy here. 24. We note that the regulatory system we have adopted for cable rate regulation on a going-forward basis is in essence a price cap scheme that is similar in many respects to the price cap regime we have adopted for the telephone industry. Rates set by applying the competitive differential are capped at that level, and cable operators may increase rates only to reflect inflation, increases in external costs such as franchise-related costs, and to reflect the costs of additional services. Our regulatory approach -- like telephone price caps -- directly regulates the prices charged to subscribers rather than indirectly regulating prices through an examination of underlying costs. Accordingly, we fully anticipate that the public will see similar benefits to those we have already witnessed from our telephone price cap regulations: lower prices for regulated services, service innovation, and increased operator efficiency, all of which contribute to industry growth and increased competition. We also believe that, as the cable and telephone industries converge, it is important to treat them with as much regulatory parity as possible. Adoption of our revised approach for regulating cable rates thus is a significant step in the right direction. 25. The specific changes we are making on reconsideration are as follows: First, as noted above, we are strengthening our statistical and economic model for estimating the difference between the rates charged by systems facing effective competition and those that do not. In our April 1993 Rate Order we estimated the competitive differential to be approximately ten percent. Numerous petitioners have challenged our methodology for deriving that figure on a variety of grounds. In response to those challenges, we have undertaken a broad review of our methodology and have refined it. 26. Our new methodology was developed in part by addressing the statistical issues identified by commenters and our staff, such as using a corrected data set and revising the treatment of equipment and installation revenues. The principal difference in calculating the new competitive differential, however, is a revised economic analysis that reflects more accurate assessments of the different types of systems that the statute defines as subject to effective competition. As noted above, Congress listed three types of systems as being subject to "effective competition" -- low penetration systems, overbuilds, and municipals. As also noted above, Congress directed the Commission, when determining whether rates are reasonable, to be guided by the goal of protecting subscribers from rates higher than those charged by those three categories of systems. 27. Statistical analysis reveals that each of the three types of systems has its own competitive differential. Because the three classes of systems differ from each other, we conclude that it is more appropriate to consider the competitive differential for each type of system individually than it is to average the data relating to all three system types. We also conclude that we are not required by the statute to compute the competitive differential simply by averaging, without evaluation, the rates charged by the three different types of systems. Rather, we interpret the Act as requiring us to "take into account" or "consider" the rates charged by each type in determining reasonable rates. 28. In selecting the best method of setting reasonable rates on reconsideration, we conducted an economic analysis that considered the competitive differential for each of the three categories of systems that Congress defined as facing "effective competition," as well as the other statutory factors, the facts of record, and the comments of interested parties. This analysis reveals that the rates of low penetration systems are not statistically different as a group from the rates of systems subject to rate regulation. We are not persuaded, however, that low penetration systems and noncompetitive systems are charging reasonable rates. Rather, we conclude that there may be a variety of reasons other than competitive pressures, as measured by the statutory definition, that have led to low penetration rates. While Congress has concluded that low penetration systems should be exempt from rate regulation, it does not follow that the rates they charge provide the best basis for determining the competitive differential that should apply to noncompetitive systems. 29. We further believe that systems in the overbuild sample provide the most informative data with regard to estimating reasonable rates. Our best estimate of the difference between the rates charged by overbuilds and noncompetitive systems is 16 percent. That figure takes into account the fact that cable operators generally do not compete head-to-head in the entire franchise area they serve. Specifically, our data show, as we expected, that rates decrease as the extent of competition increases. We thus have corrected for the lack of full competition throughout an entire franchise area when computing the competitive differential for overbuild systems. 30. In focusing on the overbuild sample, it is also appropriate to adjust the 16 percent figure to take into account the fact that cable operators serving the same area may adopt parallel or coordinated pricing practices. That is, they may recognize, over time, that it makes more economic sense not to compete vigorously, but to coordinate their prices tacitly. Economic theory states that firms operating in oligopoly structures may be expected to behave in that manner. Thus, prices observed in an overbuild situation may well be above the purely competitive level. We therefore conclude that the observed 16 percent average competitive differential for overbuild systems underestimates the true competitive differential. Our data indicate that the rates charged by overbuild systems are lowest at the outset of competition and then rise over time. This is consistent with parallel pricing behavior and strengthens our conclusion that the best estimate of the overall competitive differential is greater than the overbuild differential of 16 percent. 31. The largest differential, 37 percent, arises in the comparison of the rates charged by noncompetitive systems with those charged by municipal systems and the privately owned systems that compete with them. Arguably, this differential may be the most accurate measure of the competitive differential because government-operated entities may be presumed to charge reasonable rates and have no incentive to engage in the parallel or tacitly coordinated pricing practices discussed in the context of overbuild systems. It has been suggested, however, that municipal systems may not be earning a profit. The record evidence on this point is inconclusive. Because of these concerns, we separately examined the rates charged by the privately-owned cable systems in our sample of "municipal" systems. The competitive differential is equally large for these private systems, which suggests that the rates charged by the public and private systems in our municipal sample are reasonable. However, in view of the small number of systems in the municipal category (only eleven) and our concern that they may not be representative, we believe that we should not rely as heavily on municipals as we otherwise might in estimating the competitive differential. 32. After reviewing the data from all three types of non- regulated systems, but giving the most emphasis to the data relating to overbuilds, we have selected 17 percent as the revised competitive differential. In selecting that figure, we were guided by the 16 percent figure estimated from our data on overbuilds that measures full head-to-head competition. We moved upward from 16 percent to reflect our conclusion that cable operators in an overbuild situation are likely over time to develop a tacit understanding of rate levels that may limit the intensity of rate competition. However, we did not depart upward as far as we might have, despite the evidence relating to municipal systems, on account of concerns about the interpretation of the data in our municipal subsample and on account of our consideration of low penetration systems. Furthermore, we were guided by our belief that consumer welfare is best served by financially sound cable operators. 33. It should be emphasized that this 17 percent rate reduction is not in addition to the prior ten percent rate reduction that some operators already have applied. Rather, those operators that have already established rates based on a ten percent competitive differential will only be required to adjust rates by approximately seven percent. 34. Second, with the limited exception provided by the transition mechanism, we will require all noncompetitive systems that do not invoke the cost-of-service option to reduce their rates by the revised competitive differential. Cable systems that do not do so will be subject to refunds ordered by the franchising authority or the Commission once they become regulated. In the April 1993 Rate Order, we concluded that operators whose rates on the date of regulation are "below benchmark" should not be required to reduce those rates, despite the fact that they face no effective competition. As noted above, however, we have concluded upon further reflection that behavior reflective of market power may exist generally within the noncompetitive sector of the cable industry. For this reason, we have concluded that it is preferable to apply a single percentage adjustment to all noncompetitive systems. 35. At the same time, our lack of cost-price data regarding small operators and relatively low-priced systems persuades us that we should not apply the full competitive differential to these categories as of the effective date of our new rules. On the basis of our refined statistical analysis, we therefore continue to believe that our primary approach for regulating cable rates at this time should incorporate a benchmark mechanism. Accordingly, cable operators whose September 30, 1992 rates (when adjusted by the full 17 percent competitive differential and permitted increases) are above the revised benchmark must apply the full differential or invoke cost-of- service proceedings. Cable operators whose March 31, 1994 rates are at or below the new benchmark do not have to reduce their rates at this time, pending the Commission's collection and analysis of information about such operators' prices and costs. And, cable operators whose March 31, 1994 rates are above the new benchmark, but whose rates would be at or below the new benchmark if the revised competitive differential were applied in full to their September 30, 1992 rates, must bring their rates down to the new benchmark immediately, but do not have to apply the full competitive differential pending our analysis of prices and costs. At the conclusion of our cost study, cable operators that do not apply the full revised competitive differential immediately will be required to do so unless our analysis reveals that application of the 17 percent differential to those systems would be inappropriate. 36. We are also not requiring small operators, defined for these purposes as operators serving a total subscriber base of 15,000 or fewer subscribers, and not affiliated with or controlled by larger operators, to apply the revised competitive differential immediately. While some commenters have submitted evidence to suggest that some smaller systems may face higher costs compared to larger systems, the absence of industry-wide cost data leaves us unable to conclude that all small systems face systematically higher costs. Moreover, while publicly available financial data indicate that large operators are better able to absorb further rate reductions, the financial data we have concerning small operators is not as extensive or as reliable as the data we have concerning large operators. Accordingly, we are not requiring small operators to apply the competitive differential immediately pending our collection and analysis of information about costs and prices. As with other operators subject to transition treatment, however, small operators will be required to apply the full revised competitive differential unless our analysis of costs and prices demonstrates that the revised competitive differential should not be applied to them. 37. Third, we are providing new guidelines with respect to the regulatory treatment of "a la carte" packages. Under the 1992 Cable Act, the rates for channels offered on a stand-alone basis ("a la carte" channels) are not regulated. In our April 1993 Rate Order, we held that operators may offer subscribers discounted packages of "a la carte" channels without subjecting the package price to rate regulation as long as the channels continued to be offered on a stand-alone basis. We adopted this approach in an effort to encourage increased programming choices and discounts for consumers. We did not anticipate that many operators would use this technique to take channels out of regulated service tiers and offer them on an unregulated basis. 38. A September 1993 Commission survey (hereinafter "Competitive Survey") reveals, however, that many operators have taken that course. Specifically, of the 25 largest multiple system operators included in the Competitive Survey, 12 removed channels previously offered on regulated program service tiers and began offering them in "a la carte" packages. Moreover, the terms and conditions of some "a la carte" packages offered following rate regulation have raised issues as to whether the option of purchasing the channels on a stand-alone basis is real or illusory. 39. Accordingly, in an effort to ensure that "a la carte" offerings provide subscribers with realistic service choices and to protect against prohibited evasions of rate regulation, we are announcing interpretive guidelines for determining whether the rates for an operator's collective offering of "a la carte" channels should be regulated or nonregulated. Using these guidelines, local authorities may make initial determinations as to whether a collective offering of "a la carte" channels should be considered a regulated tier. Cable operators or consumers may then make an interlocutory appeal to the Commission of the local authority's decision. Local authorities may also request the Commission to make the initial "a la carte" decision by means of a petition for declaratory ruling. 40. Fourth, we adopt a "going-forward" methodology in this Order to adjust rates when operators add or delete channels from a regulated service tier after they have become subject to regulation. Under this methodology, the efficiencies and economies of scale that arise as operators add channels to their systems are passed on to subscribers. At the same time, operators may recover the full amount of programming expenses associated with added channels, plus a markup on new programming expenses of 7.5 percent. This methodology provides a relatively simple way for operators to adjust rates. It also provides appropriate incentives for operators to provide additional, high quality programming. The going-forward methodology accordingly will promote our goals of assuring reasonable rates while encouraging the continued growth of the cable industry. 41. Fifth, we revise our rules to provide greater administrative relief to small systems, defined in the Act as systems with 1,000 or fewer subscribers. In particular, until we develop average cost schedules for equipment, we relieve small systems owned by small operators of the requirement that they establish unbundled equipment rates based on actual cost because that requirement appears to have imposed substantial burdens on small operators. Instead, eligible systems, if they so choose, may make rate reductions by reducing each regulated billed item by 14 percent (which equals the 17 percent competitive differential reduced by approximately three percent inflation that occurred between October 1992 and September 1993). These systems are not required to file Form 1200. In addition, we will permit all operators of small systems that do file FCC Form 1200 to aggregate their equipment costs for their small systems when establishing equipment rates. Finally, we are terminating our stay of rate regulation for small systems. We will entertain requests from small systems to extend the period of time in which they must comply with our rate regulations where they can demonstrate that specified hardship conditions exist. 42. For all cable systems subject to regulation, the rates permitted between September 1, 1993 and May 15, 1994 (the effective date of these new rules), and refund liability with respect to such rates, will be determined by our initial regulations adopted on April 1, 1993. The lawfulness of rates in effect on or after May 15, 1994, and refund liability with respect to such rates, will be determined in accordance with the revised rules adopted in this Order. B. Regulation Governing Rates of Basic and Cable Programming Service Tiers 1. Background 43. After Congress deregulated most cable rates in 1984, the rates charged by many cable systems rose significantly. For example, the monthly rate for the most popular program tier of cable service rose by an average of 60.8 percent from November 30, 1986, to April 1, 1991. Monthly rates for the lowest priced basic service tier increased by 40 percent or more for 28 percent of subscribers. Although the number of channels on the average cable system also increased during the same time period, many consumers felt that the increases in their cable rates were nonetheless unreasonable. 44. Partly in response to concerns about rising cable rates, a number of economists have studied the cable industry to determine whether cable operators have significant market power. Many of these empirical studies were designed to predict the demand for cable services; others attempted to estimate the supply function for cable systems. A number of the studies were conducted when cable systems had far fewer channels than they typically do today and many were based on very limited data samples. Nonetheless, while the purposes and specific results of the studies vary depending on the particular variable estimated or projected, the time period evaluated, and the variables and forms of equations used, one result is clear: the vast majority of the studies found that cable systems have some market or monopoly power and that these systems earn, or would earn in the future, monopoly profits. 45. The conclusion that most cable systems exercise market power was embraced by Congress when it passed the Cable Television Consumer Protection and Competition Act of 1992. Specifically, Congress found that (1) "most subscribers have no opportunity to select between competing cable systems" and (2) "without the presence of another multichannel video programming distributor, a cable system faces no local competition." The result, Congress concluded, "is undue market power for the cable operator as compared to that of consumers and video programmers." 46. Congress accordingly charged the Commission with creating a regulatory scheme that will protect consumers from unreasonable cable rates until competition to or within cable services emerges and ensures that rates are set at competitive levels. Section 3 of the 1992 Cable Act provides that cable systems that face no effective competition, as that term is defined in the statute, will be subject to rate regulation by their local franchising authorities and/or this agency. Section 3 further instructs the Commission to ensure that the rates for basic service are reasonable and that in response to complaints, the rates for regulated upper tier cable programming services are not unreasonable. And, it requires that rates for regulated cable equipment used to receive the basic tier be based on actual cost. The Commission's regulations regarding basic tier rates are to "be designed to achieve the goal of protecting subscribers of any cable system that is not subject to effective competition from rates for the basic service tier that exceed the rates that would be charged for the basic service tier if such system were subject to effective competition." 47. In the Notice of Proposed Rulemaking ("Notice") in this proceeding, we tentatively identified a benchmark approach as the best means of implementing the rate regulation provisions of the 1992 Act. We then sought comment on several alternative bases for a benchmark, including a benchmark based on rates of systems subject to effective competition as defined in the 1992 Act, average rates of all cable systems, rates charged prior to the 1986 deregulation of the cable industry, or average costs. At the same time that we issued the Notice, we selected a random sample of cable systems from which we sought information concerning current prices, past prices, and system characteristics to aid in designing an appropriate rate regulation mechanism. We also obtained this information from a sample of systems appearing to be subject to effective competition as defined by the statute. 48. In our April 1, 1993, Rate Order, we concluded that the results of the Competitive Survey supported the findings of Congress that the rates for cable systems not subject to effective competition reflect pervasive market power. We further determined that we should use a rate regulation system based on the rates of systems subject to effective competition as the principal means of setting regulated rates at reasonable levels. Using the results of our Competitive Survey, we established a formula to estimate the amount by which the average per-channel rate charged by a noncompetitive system exceeded the rate that an cable system subject to effective competition with similar characteristics would charge. We called that difference the "competitive differential." 49. The Commission concluded that a competitive differential of approximately ten percent exists between the rates of systems subject to effective competition and noncompetitive systems. The Commission also concluded, however, that a competitive differential based on average behavior will overstate the appropriate adjustment for some systems. The benchmark approach we adopted in April 1993 sought to address this concern. Specifically, statistical analysis was used to develop a prediction of what price a noncompetitive system with a given set of characteristics would charge if it were a system subject to effective competition. This prediction was called the "benchmark." A noncompetitive cable system with rates at or below the benchmark for that system was deemed to have reasonable rates; rates above this predicted level were presumed to be unreasonable. Thus, systems whose rates were at or below the benchmark were not required to make any rate reductions, although their regulated rates were capped at current levels. By contrast, systems whose rates were above the benchmark were required to examine whether their rates were above or below the benchmark on September 30, 1992. If they were below the benchmark on that date, their permitted rate was deemed to be the benchmark. If they were above the benchmark on September 30, 1992, they were required to reduce their September 30, 1992 rates by 10 percent or to the benchmark, whichever reduction was less. Alternatively, a cable operator could seek to justify rates above the benchmark level through a cost-of-service showing. 50. In adopting this benchmark approach, we concluded, as we had tentatively done in the Notice, that benchmarks would "provide a simple way to ascertain on an individual system basis the extent to which rates exceed the competitive rate level," and as such, best met the statutory mandate that we reduce administrative burdens on subscribers, cable operators, franchising authorities, and the Commission. We also confirmed our tentative conclusion that while the alternative cost-of- service approach to rate regulation had its advantages, its disadvantages -- which included reducing operator incentives for efficiency and improved service, and imposing heavy administrative and compliance costs upon regulators and regulatees -- were significant. 51. In our First Recon. Order in this proceeding, we affirmed our decision to use a benchmark system based on rates charged by systems facing effective competition as the primary method of assessing the reasonableness of regulated cable rates. While several petitioners challenged our primary reliance on rates charged by systems subject to effective competition in setting benchmark rates, arguing that we failed to take into account the other statutory factors set forth in the Act, we concluded that we properly placed primary weight on rates of systems subject to effective competition in devising the benchmark approach. We also explained that while we ultimately based the rate-setting methodology on the rates and other characteristics of systems subject to effective competition, we in fact took all of the statutory factors into account when developing the entire rate regulation scheme, which, in addition to the benchmark mechanism, includes cost-of-service showings and price caps. 52. While we addressed certain issues regarding the benchmark approach in the First Recon. Order, we indicated that other issues would be addressed in a subsequent reconsideration order. Those issues include concerns about the accuracy of the random and competitive sample data used in constructing the benchmark, the methodology and statistical analysis used in developing the benchmark, and petitioners' recommendations that other variables be incorporated in the benchmark formula. 53. The analyses and suggestions of petitioners, as well as the additional theoretical and empirical analyses conducted by our staff, persuade us that the approach that we adopted in April of last year is fundamentally sound but can be refined to improve its accuracy and better meet the goals of the statute. Accordingly, we are modifying the benchmark approach in several key respects, as discussed in the following sections. 2. Impact on the National Economy 54. Before presenting our analysis of specific issues resolved in this Order, we present an analytical overview of the impact of our decisions on the Nation's economy, reviewing the likely effect of our actions on promoting innovation, investment, and growth in the cable industry. 55. We believe that the Nation will benefit from regulation that ensures the availability of cable services at competitive rates, while also ensuring that rates are not so low as to inhibit investment in new programming services and enhanced video, voice and data services in the future. 56. Any discussion of the effects of cable reregulation on investment must recognize an important principle. Investment and innovation depend on two factors: (1) incentives and (2) access to capital. Both of these factors depend critically on future rates for cable services because those future rates largely determine the returns that the investments will earn. It is useful to discuss incentives and access to capital separately. 57. We believe financial incentives to invest in the cable industry depend on the prospect of future returns, which in turn depend on future demand and how rates will be regulated on a going-forward basis. Throughout this Order, we have been mindful of taking actions to ensure that the methodology to be used in determining rates following growth in programming services, channel additions or deletions, and system upgrades encourages economically efficient investment. Moreover, by limiting the extent to which operators have an opportunity to charge rates for existing services without the constraints of competition, regulation may spur cable industry entrepreneurs to devote increased effort to innovate in ways that create net economic value. We also recognize, however, that short-term investment effects of this revised approach are uncertain and depend upon cash flow and revenue analysis by individual operators. 58. With respect to the going-forward treatment of cable services, two points are of central relevance. First, those services that will be subject to continuing regulation will be allowed to earn a competitive return. For example, operators that add new channels to regulated tiers will be allowed to recover their programming costs, including a return on investment. A second key point is that many innovative new services (i.e., those offered "a la carte" or not fitting within the traditional definition of cable television) will not be subject to rate regulation. For these services, market forces, rather than regulation, will drive investment decisions. Those key points -- reasonable returns on regulated investments and market returns on unregulated investments -- guide our analysis of how rate regulation will affect access to capital. 59. Some have argued that forcing cable operators to lower current rates will stifle investment by reducing the cash flows needed to finance this investment. However, we must balance this concern with the goal of establishing reasonable rates for regulated cable services. We also believe that, while investment could initially be adversely affected by the reductions in the cash flows generated by current regulated services, many operators will have opportunities to generate steadily increasing cash flows from unregulated services. Operators also may have access to other sources of funding for future investments. Large MSOs in this industry historically have made extensive use of debt and equity to finance their actions. Since reregulation, credit ratings for the cable operators that have issued public debt have remained stable overall and have even improved in some instances. The ability of such companies to raise investment funds in those markets is driven by the prospect of future cash flows from new investments, not simply by cash that is currently generated by past investments. As one investment analyst reported, "the potential for future cable cash flow growth hinges on introduction of new services, not raising rates for basic channels." We note that stock prices for the larger, publicly held cable operators increased significantly in 1993. 60. Of course, higher rates for today's regulated services might well reduce the cost of borrowing to finance new investments because lenders would have greater assurance of being repaid. But any such fall in borrowing costs would not represent an increase in economic efficiency. Rather, it would simply reflect the fact that the investment risk was being placed on today's consumers of regulated cable services. 61. We also observe that parties arguing the importance of current cash flows make this argument despite the fact that many of the services that will result from these investments will be unregulated. Given the important relationship between investment and unregulated services, cable systems that now offer regulated services without being subject to effective competition, as defined in the statute, will have incentives to upgrade their systems with new capabilities and to introduce enhanced functions, such as interactivity, that are not subject to rate regulation. Furthermore, the revised rate regulation provisions will help prevent cross-subsidization from subscribers of regulated cable services to subscribers of unregulated services by assuring that rates for regulated services are reasonable. 62. Regulation also should stimulate investment by companies that supply programming, customer premises equipment, and network equipment to cable operators. Because operators will be allowed to raise their rates based on the cost of new programming, program vendors will be able to charge market-driven prices for their services. As a result, consumer demand will determine the success or failure of new program offerings. Indeed, because regulation will constrain the power of cable operators relative to new regulated offerings, more of the gains from innovation may accrue to programmers, spurring their incentives to innovate. 63. Indeed, the going-forward methodology, our treatment of "a la carte" channels, our provisions for streamlined showings for upgrades and the incentive upgrade plan that we are establishing in our Cost Proceeding, will encourage operators to expand programming and service choices. Our going-forward methodology includes a 7.5 percent mark-up on costs of programming. Our "a la carte" rules remove from rate regulation certain packages of per channel offerings where consumer benefits are likely to result. The streamlined cost- of-service showing for upgrades permits operators to adjust capped rates for costs of upgrades without a review of unrelated costs. And, the incentive upgrade plan contained in our interim cost-of-service order gives cable operators substantial flexibility in setting rates for new programming and new services. 64. Equipment sales and investments also may be stimulated by the changes in this Order. Because the quantities of services demanded rise as prices fall, customer premises equipment sales, and thus investment in such equipment, could ultimately increase due to the lower prices that operators are forced to charge for regulated services. Reregulation also will provide investment incentives for the firms that provide plant and equipment to the cable operators themselves. The same forces identified above as creating incentives for cable operators to invest in unregulated markets -- the prospect of earning a market return -- could also create incentives for their equipment suppliers to make investments. 65. Reregulation also should stimulate demand for network equipment. In 1993, industry observers forecasted that construction spending would increase to $2 billion from its 1992 level of $1.3 billion. Measured in terms of percentage growth, cable operators have become the most aggressive purchasers of fiber optic equipment in the telecommunications industry. The National Cable Television Association (NCTA) recently estimated that $14 billion will be spent over the next ten years to rebuild 75 percent of the Nation's cable systems, including heavy investment in fiber optic equipment. 66. We further believe that the new and revised rules we adopt in this Order will decrease the regulatory risk faced by investors in the cable industry. By creating an uncertain environment, unstable or ever-changing regulations can discourage investment. We continue to believe that it is important to have a stable and predictable regulatory scheme and that the rules adopted today establish such a system as a transition to competition and, when appropriate, deregulation. Such regulation should also help stimulate construction of advanced networks that will become key links in the national information infrastructure. 3. Estimating the Competitive Differential 67. As noted, throughout our regulatory process we have been mindful of promoting innovation, investment and growth in the cable industry while at the same time ensuring reasonable rates for regulated services. Estimating the competitive differential requires an analysis of the difference in rates charged by noncompetitive cable systems and cable systems subject to effective competition. In this section we consider our specific methodology for calculating the average competitive differential based on the results of our Competitive Survey that we conducted as part of our initial implementation of the Cable Act. 68. In our April 1993 Rate Order, the Commission used regression analysis to compare the rates charged by cable systems that were not subject to effective competition and those Congress had defined as being subject to effective competition: i.e., low-penetration, overbuild, and municipal systems. In our Competitive Survey of September 30, 1992 cable rates, we evaluated as one group the entire competitive sample. That is, we combined all the data from the three types of systems subject to effective competition rather than evaluating each of the three types of systems separately. Using that approach, we found the average competitive differential -- the average difference in rates charged by systems subject to effective competition and those not so subject -- to be ten percent. As a practical matter, that figure gave primary weight to the data from low penetration systems because they constituted more than half of the data sample of systems subject to effective competition. 69. Questions had been raised concerning the propriety of using the low penetration systems in our analysis, and the exclusion of these systems would have had a significant effect on the competitive differential under the benchmark methodology adopted at that time. In fact, when the low-penetration data were deleted from the group and the competitive differential was estimated based on the overbuild and municipal systems data alone, the competitive differential was found to be 28 percent. We therefore issued a Further Notice with the Rate Order in which we requested comment on (1) whether the exclusion of low penetration systems would produce a better measure of the competitive differential, and (2) whether we should, and lawfully could, include within the data upon which the competitive rate differential is determined, only the rates of overbuild and municipal systems. 70. In our Second Report and Order, we concluded that we were required to consider all three types of systems subject to effective competition, as defined by Section 623(1)(1) of the Communications Act, 47 U.S.C. Section 543(1)(1), in crafting our benchmark system. Accordingly, we concluded that data from cable systems with less than 30 percent penetration should continue to be included in the sample of systems used to assess the reasonableness of cable rates. In addition, we stated that it would not serve the public interest to exclude low penetration systems merely because such an exclusion would result in larger rate reductions. We also decided to continue to include municipal and overbuild systems in the sample of systems used to assess the reasonableness of cable rates both because Congress defined those types of systems as being subject to "effective competition" and because economic analysis indicates that data from these systems are relevant to this assessment. 71. Our decision to include low penetration systems in the competitive sample used to calculate the competitive differential has been challenged on reconsideration by NYNEX. In addition, numerous other petitioners have raised questions concerning our underlying economic and statistical methodology. In response to those concerns, and based on our own further analysis, we have refined our method of computing the competitive differential. As described below, we have: (1) disaggregated our analysis of the three types of systems subject to effective competition to achieve a better understanding of the differential between these systems and those in our noncompetitive sample; (2) improved the variables used in our statistical analysis of the competitive differential; and (3) tested whether additional variables should be included in our regression analyses to account for the effects of cost and demand differences that our initial statistical estimation may not have fully captured. a. Disaggregated Treatment of Low Penetration, Overbuild, and Municipal Systems 72. The competitive differential in the Rate Order was a single number (9.4%) based on the difference between the rates of systems in the Commission's random sample of noncompetitive cable systems and the rates of low-penetration systems, overbuilds, and municipal systems. Because the competitive differential in the Rate Order was a single number calculated from the rates of all systems in the sample that were thought to be subject to effective competition, it reflected the relative number of cable systems in each of the three categories subject to effective competition. We recognize, however, that the three portions of the competitive sample -- low-penetration systems, overbuilds, and municipals -- have very different characteristics. In addition to reviewing the definitions of the three types of systems, it thus is useful to consider some of the characteristics of each. 73. Low-penetration systems are those whose subscribers comprise less than 30 percent of the households in the franchise area. The question of whether the entire franchise area was wired for cable service was not considered in defining low penetration systems for purposes of the competitive sample. Thus, this definition of "penetration" does not conform to the standard industry definition of penetration as the ratio of subscribers to homes passed by cable wire. Rather, some of the systems in our sample are considered low penetration systems even though a very high percentage of the homes passed by cable wire subscribe, since such a system may be considered a low penetration system if a large portion of the franchise area is not wired. 74. The second statutory category, overbuild systems, occurs in markets where two or more cable systems or multichannel video providers each make its service available to at least 50 percent of households in the franchise area. In order to be considered an overbuild, more than 15 percent of households in the franchise area must subscribe to services provided by other than the operator with the largest share of the subscribers within the franchise areas. To meet the statutory definition, however, it is not necessary for the cable systems actually to compete head-to-head by offering their service to all of the same potential subscribers. 75. The third statutory category, municipals, includes both cable systems owned by municipal authorities and privately-owned cable systems in franchise areas where the franchising authority itself operates a cable system offering service to at least 50 percent of the households in the franchise area. This part of our competitive sample consists of pairs of cable systems, one publicly-owned municipal system and a privately-owned system that competes with it, serving the same franchise area. All but one of these pairs meet the criteria for the "overbuild" category as well as the "municipal" category. 76. In response to petitioners' concerns regarding the validity of our statistical approach to the analysis of the competitive sample, we have conducted a more refined analysis. This revised analysis indicates that the competitive differential varies widely across the three types of systems defined as subject to effective competition. Moreover, we found these differences to be statistically significant. Accordingly, we find that instead of computing the competitive differential by analyzing collectively the rates of all three types of systems in our competitive sample, it is most appropriate to estimate a separate competitive differential for each of the three classes of systems. b. Improved Variables 77. In revising our methodology to determine reasonable cable rates, we also have improved several other aspects of our analysis. 78. First, we have refined our measure of the price of regulated cable service. Instead of measuring rates as monthly revenue per subscriber per channel, we now measure rates as monthly revenue per subscriber. Our analysis shows that the price of cable service increases only a small amount as the number of channels included in regulated tiers of service increases. Moreover, subscribers actually purchase regulated service not on a per-channel basis, but in tiers consisting of several channels. Revising the price variable in this way thus better reflects consumer demand and improves the quality of the statistical analysis. 79. In addition, we have strengthened our measure of overbuild competition. First, we have added to the overbuild sample those municipal-category cable systems that also meet the overbuild criteria. Second, because the extent of actual competition is limited in many franchise areas classified as having overbuilds, we have refined the overbuild variable to measure the extent of system-overlap competition rather than simply adopting a variable that only indicates whether or not there is any competition from another cable system. We expected and found that the effect of competition on a system's prices would be greater the larger the fraction of households in the system's service area that are subject to actual head-to-head competition. 80. The staff also detected and corrected numerous errors in the data, based on their own checks of the data and errors reported by commenters, cable systems, and users of the data. When apparent errors were found, correct information was obtained from cable systems and entered into the database. c. Tests for the Effects of Cost and Demand Differences 81. Some petitioners claim that in the original analysis of cable rates, we failed to consider important characteristics of cable systems that might have affected their rates. Controlling for such variables is important because, if competitive systems generally differ from noncompetitive systems in some characteristic that affects rates, failing to account for that characteristic in the analysis will distort the results, giving a biased estimate of the effect of competition. Suppose, for instance, that competitive systems had higher density (subscribers per mile) than noncompetitive ones, and that higher densities were associated with lower rates. In such a situation, failing to control for differences in density would cause competition to appear to reduce rates more than it actually does. Given the additional time to analyze our data on reconsideration, we have had an opportunity to control for more characteristics that may affect rates and to arrive at a more accurate estimate of the competitive differential. 82. In our recent analysis of the data we incorporated several variables that were not included in the original benchmark regressions. Among the variables used in the equation were measures of the numbers of channels of various types. We found that rates increased as (1) the number of channels increased, and (2) as the percentage of channels available only on cable (i.e., non-broadcast channels) increased. The number of local broadcast channels carried was expected to measure competition from local broadcast stations and was expected to have a negative effect on rates; however, it did not prove to be statistically significant. 83. We also considered measures of the quantities of optional services purchased by customers, including numbers of additional outlets, converter boxes, addressable converters, remote controls, tier changes, installations, and upper tiers of service. These variables, which were all measured on a per- subscriber basis, were examined to improve the measure of the competitive differential by ensuring that it reflected price differences rather than differences in the levels of service consumed. Increases in the percentages of customers purchasing remote controls, additional outlets, upper tiers of service, and tier changes all were associated with increases in revenues per subscriber. 84. Other variables were considered to reflect costs of providing service, including the number of subscribers to the system. As expected, rates per subscriber declined as the number of subscribers increased because fixed costs could be spread over more customers. Construction costs were represented by various measures of miles of plant (i.e., cable wiring), miles below ground, miles of fiber plant, and whether a system was required to bury cable drops. These variables were not statistically significant and were dropped from the analysis. At several points in the analysis, various measures of density were tested and never proved to be statistically significant. Whether a system was owned by a multiple system operator (MSO) was included in the expectation that economies of joint ownership might exist. Surprisingly, systems that were owned by an MSO had higher per- subscriber revenues than those that were not. In addition, we expected that high wage rates would raise system costs. Median income in the franchise area was used as a proxy for wage rates, and median income was positively related to revenue per subscriber. 85. Contrary to claims of some petitioners, we also do not believe that the benchmark is flawed because it does not account for differences in individual system costs. The regression analysis performed has identified the statistically significant determinants of average revenue per subscriber. Some of these are cost-related, some may reflect demand factors, and some may embody both demand and cost factors. The data did not allow us to isolate a statistically significant relationship between costs and prices so as to permit us to craft a cost-based benchmark. Cable systems retain the option to initiate a cost- of-service proceeding if they believe that the benchmark fails to provide them with a reasonable return. We also note that in our Cost Proceeding, we are examining the use of average cost schedules to set regulated cable rates which should provide an approach to setting rates that is similar to a cost-based benchmark. d. System Size and the Competitive Differential. 86 Several petitioners conducted analyses of the Commission's cable Competitive Survey data that could be interpreted as demonstrating that the competitive differential varies with system size and that the competitive differential is not statistically significant for systems larger than 5,000 subscribers. These commenters failed to offer an explanation of why it is appropriate to break up the sample at this point, why such an effect might be expected to occur, or why we should take it into account in formulating policy. 87. We believe that to some extent the failure to find a significant competitive differential in the analyses of a reduced sample comprised solely of larger systems stems from inaccurate measurement of the degree of competition. First, inclusion of low penetration systems in the competitive sample, even though they appear to behave no differently from noncompetitive systems, caused the estimate of the competitive effect to be smaller and less statistically significant for systems of all sizes. In other words, were it not for the disproportionately high number of low penetrations systems in the competitive sample, the competitive differential would have been larger for systems of all sizes. 88. Second, the apparent lack of a significant competitive difference in the rates of larger systems may have resulted from measuring competition on a franchise area basis when many systems charge uniform prices for all franchise areas in the system. Using the system overlap measure of competition, described above, which we believe more accurately captures the extent of competitive pressure on prices, we tested for the extent to which prices varied with system size for competitive and noncompetitive systems. And we used a procedure that does not arbitrarily divide the sample at 5,000 subscribers. Using this procedure, we found that the effect of system size on the competitive differential is not significant. This finding supports the conclusion that a single competitive differential is appropriate for systems of all sizes. 89. We also considered another statistical approach that yielded different competitive differentials for systems of differing sizes. This approach leads to the conclusion that the rates charged by cable operators that are subject to competition rise to the level of noncompetitive ones as the number of subscribers increases. To draw policy conclusions, one must understand the source of any relationship between system size and the magnitude of the competitive differential. One possible explanation for why the competitive differential may be smaller for larger systems is that larger competitive systems are more sophisticated and may thus have learned to collude more effectively. To the extent that this explanation is valid, it suggests that the larger competitive differential exhibited by smaller systems more accurately represents the effect of competition. This logic would suggest that we apply the larger competitive differential derived from smaller systems to all systems, regardless of their size. Nevertheless, given the relatively small number of observations in the competitive sample, we believe that dividing the data into smaller and smaller samples is statistically risky, and that a competitive differential estimated on the basis of the three types of systems subject to effective competition is more likely to yield an accurate measure of competition. e. The New Competitive Differential 90. Our statistical analysis of the Competitive Survey indicates that the competitive differential is 1 percent for the low penetration sample, 16 percent for the overbuild sample, and 37 percent for the municipal sample. These three numbers serve as the fundamental components of the Commission's derivation of an appropriate composite competitive differential for application to the rates of regulated systems. In deriving a single competitive differential, the Commission is compelled to exercise its judgment and expertise, taking into account the factors identified by Congress in the Cable Act. In so doing, the Commission is especially guided by (1) the Act's requirement that in promulgating rate regulations the Commission is to take into account or consider, inter alia, the rates charged by systems subject to effective competition and (2) Congress's finding that "without the presence of another multichannel video programming distributor, a cable system faces no local competition," resulting in "undue market power" for the cable operator. 91. In determining a single composite competitive differential to apply to regulated systems, we believe the Cable Act of 1992 requires us to take into account all three categories of cable systems that Congress defined as being subject to effective competition. For that reason, we deny NYNEX's petition for reconsideration of our Second Report and Order. That petition urges us to exclude low penetration systems from the competitive sample. In considering the rates of the three types of systems, however, we believe it is appropriate and informative to assess the extent and nature of competition faced by each of the three classes of systems that Congress has deemed to be subject to effective competition. 92. As noted, low penetration systems' rates did not differ substantially from the rates of systems in the random sample not subject to effective competition. While low penetration may result from sharing a market with a competitor, it may also result from a number of other conditions not related to competition. For example, a new system that has just begun operation or a system that serves a low income neighborhood may have low penetration. A system with high prices or poor service also may have low franchise area penetration as a consequence. 93. In conducting our data analysis, we made a number of attempts to isolate these various factors and their potential effects. Regressions were run to examine the effects of consumer income levels, the extent to which a system serves an urban population (in an attempt to capture the effects of local competition from other forms of entertainment), the age of the system, and the number of broadcast television channels in the service area. None of these attempts was successful in identifying a statistically significant effect, leaving us unable to conclude that the rates set by low-penetration systems (other than those that also are overbuilds or municipal systems) are likely at or near competitive levels. 94. Furthermore, the statutory definition of low penetration is based on homes in the franchise area, not homes passed by the cable system. Yet cable systems frequently offer service to only a portion of the franchise area. Indeed, it is possible that, under the statutory definition, a system could have low penetration simply because it is redlining (i.e., choosing to offer service in only part of its franchise area). When measuring the choices available to potential subscribers, penetration might better be measured on the basis of homes passed rather than franchise area homes. Hazlett has examined the franchises in the low penetration sample and found that a significant minority of them in fact exhibit greater than 30 percent penetration of homes passed. Our own data analysis reaches a similar conclusion. 95. Previous studies of cable industry market power conducted by outside experts and academics have focused on overbuilds as representing competition because they are engaged in head-to-head competition. We agree that overbuilds come closest to facing competition and thus charging reasonable rates, although interpretation of the overbuild results must be guided by analysis of the other two types of system deemed by Congress to face effective competition. 96. There were 41 overbuild systems in our sample -- 51 when the overbuild municipals are included. As noted above, it is not necessary for overbuild systems to actually compete head- to-head across their service areas in order for them to qualify as overbuilds under the statutory definition. However, our analysis of the Competitive Survey data reveals that there is extensive head-to-head competition in many instances. While this situation is closer to the concept of robust and vigorous competition than is the situation of a low-penetration system, we still believe that we must also consider other factors that will reflect aspects of a competitive marketplace. 97. As explained more fully in the Technical Appendix, we corrected statistically for the fact that most overbuilds do not entail full head-to-head competition. We were able to do so because the amount of head-to-head competition varies among the overbuilds in our sample and the rates charged by cable companies decrease as the amount of head-to-head competition increases. Accordingly, we calculated a revised competitive differential of 16 percent for the overbuild systems in our sample. We regard that figure as superior to the 12 percent figure derived from the data without correction for the lack of full head-to-head competition. 98. Many petitioners agree that the rates charged by overbuild cable systems reflect competitive behavior. Indeed, petitioners also urge us to rely primarily on these rates for purposes of determining the competitive differential. There are, however, two concerns with the use of the overbuild competitive differential as the measure of the overall average competitive differential. One point argued by some petitioners is that overbuild operators may set prices below their average total costs (i.e., below the costs that include a return on capital), particularly during the early stages of overbuild competition. Petitioners suggest, for example, that the competitive rivalry between two cable competitors in the same franchise area leads to unsustainable price cuts or "price wars," during which services are priced below cost. William Shew, on behalf of Harron Communications, argues specifically that overbuilds that have been in existence for less than five years are not a reliable indicator of competitive prices because they engage in price wars. As evidence of this, Shew states that the overbuilds in the FCC database that have been in existence for less than five years charged rates that are 25 percent lower than those overbuilds where competition endured for more than five years. Shew argues that rates charged by overbuilds in existence more than five years are reliable measures of competitive rates because prices have stabilized. 99. As we stated in the Second Report and Order, there is nothing in the record to support the contentions that overbuild systems of any particular age are charging rates that do not allow them to recover costs or otherwise provide for viable operation. Even if Shew is correct that new overbuilds charge lower rates than more mature overbuilds, he presents no evidence that the newer systems are failing to earn a profit or will be unprofitable in the long-run. Equally important, the behavior cited by Shew is equally consistent with the well-established body of economic evidence showing that firms that face one or at most a few competitors may eventually collude and collectively raise prices rather than compete with each other. 100. Thus, the second concern about the overbuild competitive differential is that it is smaller than one that would be generated by vigorous competition. Several market conditions indicate that cable overbuilds would be expected to charge parallel or coordinated rates successfully. First, there typically are only two cable operators in a given overbuild franchise area. Second, operators can observe one another's marketing efforts and thus respond quickly to deviations from an implicit agreement. Third, operators do not enter into long-term contracts with their subscribers, so that if one operator deviates from implicit parallel rates, the other operator can retaliate to take the customers back, thereby lowering the profitability of a practice that would deviate from the coordinated, parallel rates. Finally, cable operators are not dealing with large, organized customers who might otherwise be expected to exercise buyer power. 101. Both the "price-war" and the "parallel or coordinated pricing" scenarios predict that competitive differentials should fall as the length of time during which the systems have been in competition with one another rises. Statistical analysis of the data show that the competitive differential falls over time. In particular, various exploratory regressions show that operators in young overbuild situations price as much as 25 percent less than operators in the noncompetitive sample, while systems that have been in competition five years have a competitive differential of approximately 17 percent, and by some measures the competitive differential falls to ten percent or less after an even greater number of years. While they are informative, no one number is definitive because each is derived by taking the overbuild subsample and dividing it further. From this regression analysis and our examination of the industry structure, we believe that estimates of ten and 12 percent for the competitive differentials are clearly too low due to the effects of parallel or coordinated pricing, and the true average competitive differential is substantially higher. At the same time, this analysis leads us to conclude that these effects are unlikely to push the true differential much higher than 20 percent. Consequently, our best estimate of the average competitive differential starts with the overbuild coefficient from the benchmark regression of 16 percent and is adjusted slightly upwards. 102. The municipal sample clearly demonstrates the lowest rates, with an estimated competitive differential of 37 percent. Our sample of municipal systems consists of 11 systems, 10 of which also qualify as overbuilds. In light of the possibility of tacit coordination between oligopolists discussed above, and the likelihood that government authorities are less inclined to participate in such practices, the systems in the municipal subsample might be thought to present the best measure of competition. 103. However, we find there are several concerns with the municipal sample. First, the municipal sample contains only 12 operators. We raise this concern both in the statistical sense of using a small sample, and because it demonstrates that the universe of municipal systems is itself small and thus, in some respects, unusual as compared to other competitive systems. Petitioners have also raised concerns that the very low rates charged by municipally-owned systems reflect "unfair" competition. They suggest that municipalities have lower costs than privately-owned systems, are willing to accept losses, and may be subsidizing the provision of cable service from other revenues. The National Cable Television Association (NCTA) attempted to support these contentions with studies showing that a few municipal systems were subsidized or not making a profit. However, these municipal systems responded that the information and assumptions used in the studies were not accurate and that NCTA's conclusions were unfounded. 104. Because of these concerns about municipally owned systems, we turned our attention to the private systems that compete against municipally run systems. Analysis revealed that the competitive differentials for private and municipal systems do not differ from one another by a statistically significant amount. In every instance, however, the municipal system was built after the private system had been operated for some portion of time, and several of the municipals appear to be quite new. Thus, we cannot be sure that the privately operated systems are continuing to cover their investment costs at the current prices. 105. After reviewing the disaggregated data from all three types of unregulated systems, we have decided that we should give the most emphasis to the data relating to overbuilds. After doing so, we have selected 17 percent as the revised competitive differential. In selecting that figure, we were guided by the 16 percent competitive differential estimated from our data on overbuilds to reflect full head-to-head competition. We adjusted upward from 16 percent on account of our conclusion that cable operators facing competition may engage in parallel or coordinated pricing over time. We also considered the 37 percent competitive differential for municipal systems but discounted this factor somewhat because of the small number of municipal systems and on account of our consideration of low penetration systems, which had only a one percent competitive differential. Our decision to adjust upward was also influenced by our conclusion that any cable operators that would be harmed by applying the competitive differential because analysis of their costs and revenues shows that they do not fully exercise their market power may invoke our cost-of-service rules. 4. Applying the Competitive Differential 106. The April 1993 Rate Order required some, but not all, noncompetitive cable operators to lower their rates to avoid refund liability. In particular, only those regulated operators with very high rates were required to come down ten percent. Operators with rates less than ten percent above the benchmark were required to reduce their rates only to the benchmark which was the average per-channel rate charged by similar effectively competitive systems in our rate sample. Those cable operators with rates below the benchmark were not required to reduce their rates at all. This approach reflected concern that an average adjustment factor applied to all regulated cable systems would be too high for some systems (i.e., those for which the differences between revenues and costs were relatively small). Because the approach focused solely on the revenue part of the difference between revenues and costs, an implicit assumption of this approach was that all cable operators' costs are similar, so that only high subscriber rates reflect the exercise of market power. Expressed somewhat differently, this approach gave weight to the argument that only those cable operators with high rates were charging subscribers too much. 107. We believe that profit-maximizing cable operators, whether they are subject to competition or are noncompetitive, set their prices based on costs and subscriber demand. Without more detailed information on the cost and demand conditions facing a particular cable operator, however, it is impossible to determine the amount by which that operator's revenues exceed its costs. A cable system may have a small degree of market power yet still charge a high price because its costs are high. Likewise, a cable system may have a large degree of market power yet still charge a low price because its costs are relatively low. Given the absence of industry-wide data, we have not been able to identify the underlying cost and demand factors with sufficient precision to allow us to construct an estimate of market power on a system-by-system basis. Furthermore, we believe that our refined statistical analysis indicates that behavior reflective of market power may exist generally within the noncompetitive sector of the cable industry, rather than just with systems charging relatively higher rates. 108. This finding is consistent with the numerous other studies performed over the years. We have reviewed many studies of market power of cable systems conducted by the Commission and independent market analysts. Based on these studies, there seems to be relatively little disagreement among economists that noncompetitive cable companies possess market power. 109. For these reasons, we have concluded that it is preferable to apply the same percentage adjustment to all regulated cable systems rather than to attempt to assign different adjustments to different systems. Thus, as a general matter, to avoid refund liability, regulated cable systems will be required by May 15, 1994 (the effective date of the modified rules) either to set their rates so that their regulated revenues per subscriber do not exceed September 30, 1992 levels reduced by the revised competitive differential of 17 percent (with certain adjustments described below), or to submit a cost-of- service showing supporting higher rates. At the same time, our analysis persuades us that there are two limited classes of operators that should not be required to adjust their rates in this manner until we develop a better picture of the price/cost profiles of these operators. The requirement to set rates according to the general rule thus will not apply to regulated cable systems that fall into one of two categories of systems eligible for transition relief. 110. The first category eligible for transition relief consists of systems owned by "small operators," a term defined below as cable operators which have a total subscriber base of 15,000 or fewer customers and which are not affiliated with a larger operator. The second category consists of (i) systems whose March 31, 1994 rates are at below the revised benchmark and (ii) systems whose March 31, 1994 rates are above the benchmark but whose permitted rates are at or below the benchmark. 111. Systems eligible for transition relief will not be required to make the full reduction otherwise required until the Commission has collected and analyzed data about such operators' prices and costs, and determined whether such a reduction is not inappropriate. We anticipate that some of these price/cost data will be collected through a cost study to be conducted by Commission staff over the next six to nine months, while other data will be submitted to the agency by the affected cable systems in a proceeding on this issue to be initiated shortly. The relevant price/cost data then will be aggregated and analyzed so that they can be applied on an industry-wide, rather than a system-by-system, basis. At the conclusion of our analysis, systems eligible for transition relief will be required to make the full reduction unless the analysis reveals that application of the 17 percent competitive differential to these systems is inappropriate. 112. For all cable systems subject to regulation, the rates permitted for the period from September 1, 1993 until May 15, 1994 (the effective date of these new rules), and refund liability with respect to such rates, will be determined by our initial rate regulations adopted on April 1, 1993. The lawfulness of rates in effect on or after May 15, 1994, and refund liability with respect to such rates, will be determined in accordance with the new rules adopted herein. 113. The specifics of how the new rules will be applied, and how cable systems and regulators should transition from the old rules to the new rules, is discussed in detail in the following sections. In addition, the specific calculations a regulated cable system will need to use to apply the revised benchmark system are set forth in new FCC Form 1200. a. Systems Not Entitled to Transition Relief 114. As noted above, regulated cable systems that are not entitled to transition relief will be required to set their rates at a level that equals their September 30, 1992 regulated revenues per subscriber reduced by the revised 17 percent competitive differential and floated forward for certain adjustments. As under the old benchmark system, we are using September 30, 1992 regulated rate levels as our starting point to avoid building into permitted rate levels any unwarranted rate increases that may have occurred after the 1992 Cable Act was passed but before we adopted our initial rate regulations on April 1, 1993. Applying the revised 17 percent competitive differential to September 30, 1992 rates thus will best ensure that subscribers pay "reasonable" regulated cable rates. 115. As we previously determined, however, a system may have incurred costs since September 30, 1992 that should be reflected in its lawful rate. In particular, we believe that regulated systems should be allowed to include in their permitted regulated rates: (1) the inflation occurring between October 1, 1992 and September 30, 1993; (2) changes in external costs that have occurred since the system became subject to initial regulation at either the local or federal level (or February 28, 1994, whichever was earlier); and (3) changes that have resulted from the addition or deletion of program channels to regulated service tiers since September 30, 1992. Accordingly, after reducing its regulated September 30, 1992 rate levels by the 17 percent competitive differential, a regulated system will be allowed to adjust that rate forward for the described changes. The resulting rate will be known as the "full reduction rate." 116. A system whose rate level being justified is above its full reduction rate level must reduce its rate to the full reduction rate level, unless it qualifies for transition treatment as discussed in the sections below. By contrast, a system whose rate level being justified is below the full reduction rate will be permitted to raise its rate level up to the full reduction rate level. This is because the full reduction level establishes the reasonable rate level for that system under our rate regulations. Any cable system that sets its rates at the full reduction rate level will be entitled to adjust those rates in the future for annual inflation, changes in external costs, and changes in the number of regulated channels. b. Systems Entitled to Transition Relief (i) Systems Owned by Small Operators 117. As noted above, systems owned by "small operators" will not be required to reduce rates to the full reduction level immediately. Instead, they will be allowed to cap their rates at their March 31, 1994 levels until we have completed our study of the prices and costs experienced by systems of this type. 118. We adopt this transition approach for cable systems owned by small operators for several reasons. First, evidence submitted by petitioners in this proceeding suggests that smaller systems may face higher than average costs. This evidence is insufficient to allow us to conclude that all small systems face systematically higher costs due to the absence of industry-wide cost data. The information in the record, nonetheless, raises a legitimate question as to whether some systems (and operators) with a limited subscriber base do in fact have unusually high costs (and thus lower-than-average margins). In addition, we are concerned that some small operators may not have the financial wherewithal to withstand the impact of a significant rate reduction. We therefore believe that it is appropriate to study the costs of small operators, and compare those costs with the prices they charge for regulated services and equipment, before requiring them to reduce their rates to the full reduction level. 119. We note, however, that the existing record evidence does not enable us to find that small operators should not eventually be required to apply the full revised competitive differential. Thus, as with other operators subject to transition relief, systems owned by small operators will be required to apply the full 17 percent competitive differential unless the price/cost data we collect persuade us that a smaller competitive differential should be applied to them. 120. For purposes of deciding eligibility for transition relief, systems owned by "small operators" are defined as systems which are owned by operators with a total subscriber base of 15,000 or less as of March 31, 1994, and which are not affiliated with or controlled by larger operators. Our survey of industry rates as of September 30, 1992, the Competitive Survey, revealed that, on average, cable systems charge roughly $20 to $25 each month for all regulated services and equipment. Using this figure to estimate regulated revenue per subscriber, this means that an operator with a total subscriber base of 15,000 earns approximately $3.6 to $4.5 million annually from regulated cable service. We believe that operators who exceed this revenue level are sufficiently large that they will likely be able to apply for bank loans, credit lines or other sources of financing in their communities should application of the full 17 percent competitive differential pose financial difficulties for them. Moreover, as noted above, we have some record evidence that raises the possibility that some small systems serving 1,000 or fewer subscribers may have higher than average costs. Should this prove to be true, we do not believe that these higher costs necessarily will be experienced only by small systems serving 1,000 or fewer subscribers; some systems with slightly larger subscriber bases may also face such costs. Establishing a 15,000 total subscriber base cut-off will thus serve the additional purpose of providing transition relief to a number of smaller systems serving more than 1,000 subscribers that could possibly be found to face unusually high costs. 121. In adopting transition relief for small operators, we need to address two possible complicating scenarios. The first involves a small operator who subsequently purchases, or is purchased by, another cable operator so that the combined subscriber base of the two operators exceeds 15,000. We believe that the small operator in such a situation should not be required to forfeit its transition treatment simply because an acquisition has occurred. We accordingly will grandfather the rate treatment of the small operator pending completion of our cost analysis. We note, however, that the grandfathered treatment will apply only to the systems originally owned by the small operator, and will not extend to the new systems it has acquired (or with which it has been merged). In this way, we will avoid creating any artificial regulatory incentives that either encourage or discourage the acquisition of cable systems by small or large operators. 122. The second scenario that could potentially arise concerns cable operators who cross the 15,000 subscriber demarcation line by either adding or dropping subscribers as their business naturally expands or contracts. A system owned by a small operator on March 31, 1994 that is entitled to transition treatment will not lose its eligibility for that treatment simply because the operator grows above the 15,000 subscriber limit prior to the completion of our cost analysis. Similarly, an operator that exceeds the 15,000 subscriber cut-off on March 31, 1994 will not gain eligibility for transition relief if it subsequently loses sufficient subscribers to bring it below the 15,000 subscriber limit. We believe that these principles are necessary to avoid introducing regulatory incentives for operators either to add or drop subscribers in order take advantage of a particular rate treatment. (ii) Low-Price Systems 123. The second class of regulated cable systems entitled to transition treatment are those whose March 31, 1994 rates are below the revised benchmark, and those whose March 31, 1994 rates are above the revised benchmark but whose full reduction rates are below the revised benchmark. These systems are charging comparatively low prices when measured against other noncompetitive systems, as indicated by their position relative to the new benchmark. As noted above, we have no conclusive evidence demonstrating that these "low-price" noncompetitive systems are not exercising the market power reflected in the 17 percent competitive differential that we found to exist, on average, between the rates charged by noncompetitive systems and systems subject to effective competition. In particular, we tested these low-price systems for various demand and cost characteristics and did not find that their rates significantly varied in any systematic way from those of all other noncompetitive systems. 124. Nonetheless, we remain concerned that, because their prices are significantly lower than those charged by most noncompetitive systems, systems in this second class may face unusual demand, cost or other influences that have not been captured in our analysis to date. Accordingly, to study this issue further, we will grant transition treatment to cable systems with relatively low prices (i.e., those with rate levels below the revised benchmark and those with rate levels above the revised benchmark but whose full reduction rates are below the benchmark). We note, however, that the record evidence to date leaves us unable to conclude that these systems should ultimately be exempted from the requirement to apply the full revised competitive differential. Thus, as with small operators subject to transition relief, systems with relatively low prices will be required to apply the full 17 percent competitive differential if additional analysis of their costs fails to demonstrate that a smaller competitive differential should be applied to them. 125. In order to determine whether they are "low- price" systems entitled to transition relief, all systems that do not qualify for transition treatment under our "small operator" definition will be required to compare their March 31, 1994 rates to the new benchmark and to their full reduction rates using FCC Form 1200. Systems whose March 31, 1994 rates are below the revised benchmark, or whose March 31, 1994 rates are above the revised benchmark but whose full reduction rate is below the revised benchmark, will be eligible for transition treatment. 126. To compare its rates to the new benchmark, a cable system will first calculate its "regulated revenue per subscriber" for the franchise area at issue as of March 31, 1994. The "regulated revenue per subscriber" is the cable system's revenue from its basic and cable programming service tiers, plus its regulated equipment revenue, divided by its number of subscribers. The cable system will then calculate its "benchmark regulated revenue per subscriber" using the revised benchmark formula. This benchmark rate will be based on the system's characteristics as of March 31, 1994. The benchmark rate will incorporate the 17 percent competitive differential and will be adjusted for inflation to enable the cable system to make a proper comparison between its March 31, 1994 rates and the benchmark. c. Application of Transition Relief 127. Regulated cable systems eligible for transition relief, either because they are owned by small operators or because they are low-price, will not be required to adjust their rates to the full reduction rate level pending completion of the Commission's price/cost analysis. Rather, systems that are owned by small operators and systems whose March 31, 1994 rates are below the revised benchmark will not have to make any reductions at this time. Systems whose March 31, 1994 rates are above the revised benchmark but whose full reduction rates are below the revised benchmark will only be required to reduce those rate levels to, and not below, the revised benchmark during the transition period. We are not requiring these latter systems to reduce their rate levels below the revised benchmark for two related reasons. First, as noted above, we are studying further the question of whether below-benchmark rates are more likely to be reasonable than above-benchmark rates, because they are comparatively lower. In light of this outstanding inquiry, we do not believe it would be appropriate to require regulated systems to reduce their rates below the benchmark level at this time. Second, requiring any systems whose rates are currently slightly above the benchmark to reduce their rate levels to the full reduction levels, but not requiring below-benchmark systems to reduce their rates at all, would result in inequitable treatment of systems that may be fairly similarly situated. 128. For purposes of applying the new rate rules, a system's March 31, 1994 rate is the rate that the system was permitted to charge under the old benchmark system, which in turn would consist of its initial permitted rate plus any external costs that it accrued up to March 31, 1994. We realize that some systems have already become subject to regulation at the local or federal level, and some have not. If, on March 31, 1994, a system is involved in a pending rate proceeding before either its local franchising authority or the FCC, its March 31, 1994 rate will be the rate that the regulator ultimately decides is reasonable. It will also be subject to refund liability for the period during which its March 31, 1994 rate may have been unlawfully high as measured under our current rules. We impose this requirement because we do not intend for our transition approach to serve as a shield to rate reductions that may be required as a result of pending proceedings initiated under our prior rules. Similarly, if a system entitled to transition relief that is not already regulated becomes subject to regulation after March 31, 1994, it will be required to revise its rate and issue any relevant refunds should the regulator determine that its March 31, 1994 rates were unlawful under our initial benchmark rules. This approach is necessary both to ensure that subscribers get the benefits of rate reductions to which they would have been entitled under the old rules, and to ensure that cable systems that became subject to regulation before May 15, 1994, the effective date of the new rules, are not treated differently (and more adversely) than systems that become subject to regulation after that date. 129. We also note that our decision to relieve systems entitled to transition treatment from immediate application of the full competitive differential does not relieve those systems of our other requirements concerning the restructuring of equipment and program service offerings. Thus, all regulated systems except those excused by specific provisions in our rules remain required to (1) set equipment rates at cost (including a reasonable profit), (2) unbundle equipment charges from programming rates, and (3) apply an average rate per channel when setting program tier charges. 130. Systems eligible for transition relief will be subject to a modified price cap pending completion of the Commission's price/cost analysis. Specifically, these systems, like all other regulated systems, will have to compute their full reduction rate. They will also have to calculate their "transition rate," which is simply the rate that they are permitted to charge at the beginning of their transition period. Systems entitled to transition treatment may increase their rates to reflect increases in external costs and increases caused by channel changes that accrue after March 31, 1994. We are permitting these particular rate increases because we wish to preserve sufficient incentives for systems to retain, and add, programming channels to regulated program services during the transition period, and both external costs and the "going-forward" methodology allow recovery of program- related expenses. 131. We will not allow such systems, however, to increase their transition rates due to increases in inflation until the transition rate equals their full reduction rate. Under the revised rules, a system's full reduction rate -- which, unlike its transition rate, rises with inflation as well as with changes in external cost and channel changes -- may eventually exceed the transition rate. At the point when the transition rate and the full reduction rate become equal (if such a point occurs during the transition period), the system will be entitled to adjust its rate upward to take advantage of all future inflation adjustments. This process protects the cash flows of systems subject to transition relief, while gradually bringing their rates in line with the full reduction amounts. This smooth transition to full reduction rates should thus reduce the administrative and financial burdens of implementing the full competitive differential. d. Regulated Rates at the End of the Transition Period 132. In the near future, the Commission will initiate an industry cost study pursuant to our cost-of-service rulemaking proceeding. Information about the prices and costs of small operators and low-priced systems will be collected as part of that effort. In addition, we will shortly issue a further notice in this proceeding to enable these systems to submit additional evidence to us concerning their prices and costs. 133. The Commission will use the information it receives in these proceedings to determine what competitive differentials ultimately are appropriate for the two classes of systems eligible for transition relief. The competitive differentials for each category will then be applied on a classwide basis. In no instance will either class be subject to a competitive differential greater than 17 percent. Subject to that limitation, we expect to apply the largest competitive differential that is consistent with the average operator in each class earning the 11.25 percent rate of return that we are today adopting in our interim cost-of-service rules. 134. Once the appropriate competitive differential has been applied to a system, that system will be entitled to an "aggregate inflation adjustment" equal to all GNP-PI inflation adjustments for the period beginning October 1, 1992 through the most recent June 30. To the extent a system has already received some inflation adjustment for that period, the system will receive the net of the aggregate inflation adjustment minus any inflation adjustment already received. In either case, after the end of the transition period, a system will be eligible for additional inflation adjustments on an annual basis, but no earlier than September 30 of each year, when the final GNP- PI through June 30 of each year is released. e. Calculation of Refund Liability 135. In general, regulated systems who select the benchmark approach to setting rates will be required to comply with the revised rules by May 15, 1994 in order to avoid refund liability. We recognize, however, that there may be some systems who will not be able to bring their regulated rates into compliance with the modified benchmark system by that time. Specifically, to comply with the new rules, operators will need to collect the necessary rate- setting information, complete the applicable FCC Forms to determine their new permitted rates, and give the required notice of any rate changes to their subscribers. Although the Commission is making every effort to ensure that this process runs as smoothly as possible (by, for example, making FCC Form 1200 and related forms available on a computer disk), it is aware that some systems may have difficulty completing the necessary tasks before the effective date of the new rules. 136. Accordingly, to reduce the burden on cable systems that cannot conform their regulated rates to the new benchmark approach by the effective date of the revised rules, we will not impose refund liability on such systems for an additional 60 days after May 15, 1994 (i.e., until July 14, 1994), as long as certain conditions are met. First, systems wishing to take advantage of this deferral of refund liability may not change any rate for regulated service or equipment, or restructure any regulated service or equipment offering (by, for example, removing program channels from what would be regulated service tiers and placing them into an "a la carte" package), during the period that runs from the release date of this Order to July 14, 1994. Moreover, a cable system that does restructure its rates and service offerings, even in compliance with our rules, before July 14, 1994 will have its refund liability triggered on the date the restructuring occurs. These conditions are needed to ensure that cable systems do not attempt to game the benchmark process before the new rules take effect and during the refund deferral period. 137. Second, cable systems taking advantage of the refund deferral period must still give at least 30 days notice to subscribers of any rate or service changes they ultimately make in response to the new rules, as required under our revised notification provisions. Our experience with the rate changes that occurred shortly before the initial rate regulations went into effect on September 1, 1993 revealed that many subscribers were surprised by the rate and service offering changes that were made because we had preempted all notification requirements. We believe that consumers are better served if they have sufficient warning that their rates and service offerings may change again, even if those changes primarily result in rate decreases. 138. Third, all rate and service restructuring must be completed by July 14, 1994 (the end of the 60-day deferral period) in order to avoid refund liability. We believe that an additional two months beyond the effective date of our new rules provides cable systems adequate time in which to familiarize themselves with the regulations and take the necessary actions to comply. This is particularly true given that rate regulation is already well under way for many operators. f. Notice to Subscribers 139. The Commission's existing rate regulations require that cable systems give 30 days notice to franchising authorities of any rate increase. Our customer service standards provide that systems must give 30 days notice to subscribers before implementing any rate or service changes. These standards, however, are enforced in the first instance at the local level by local franchising authorities, and not primarily by the Commission. In order to better ensure that consumers have sufficient notice of rate and service changes and to clarify operators' notification requirements, we are modifying our rate regulations to require that cable systems give 30 days notice to both subscribers and franchising authorities before implementing any rate or service changes. 140. In addition, our experience with the rate changes that occurred in the wake of rate regulation suggests that a number of cable subscribers have either not been informed or have been confused about the reasons for recent rate or service changes. To ameliorate this situation, we are expanding the notice requirement to ensure that cable operators provide consumers with better information about why rates or services are changing. In particular, cable systems will have to identify on subscriber bills the precise amount of any rate change and briefly explain its cause (e.g., inflation, changes in external costs or the addition/deletion of channels (identified by name)). This information must be presented in a way that enables the average subscriber to understand readily why his or her rates have increased or decreased. 141. In our April, 1993 Rate Order, we did not require that cable systems provide subscribers with information regarding how to lodge complaints about rate changes, but instead only required that notices of upcoming rate changes include the name and address of the system's local franchising authority. We also did not require that rate change notices include the address of the Commission. Upon further consideration, we believe that it would impose little or no additional burden upon cable systems but would better serve consumers if systems were required to notify subscribers of their right to file complaints about changes in cable service tier rates and cable programming services with the Commission within 45 days of the change being reflected in their bill, and to provide the address and phone number of the local franchising authority and the Commission. We believe these changes in our notice requirements will make it easier for subscribers to ask the Commission to investigate potentially unjustified changes in rates for cable programming services. In order to minimize burdens on cable operators, however, we also reaffirm our prior conclusion that the notice will not have to include a copy of our complaint form. 142. These notice requirements are effective immediately upon publication of the revised rules in the Federal Register. We find good cause to make these requirements effective on less than 30 days notice in the Federal Register. 143. In addition, we are requiring that all cable system bills to subscribers contain the address and phone number of the local franchising authority and the Commission. We believe that this requirement will impose little or no additional burden upon cable systems and will be useful to consumers. g. Procedural Issues for Franchising Authorities 144. We realize that local franchising authorities and cable operators currently involved in proceedings concerning basic cable rates may have questions regarding the procedural deadlines that they must follow, given the adoption of our revised rate rules. As explained below, with one exception, we see no reason why franchising authorities and cable operators should deviate from the timeframes already established in Sections 76.930 and 76.933 of our Rules for proceedings that were initiated before the effective date of this Order. Thus, adoption of this Order generally does not affect the basic deadlines to which local authorities and operators must adhere for resolving pending rate cases under our initial benchmark regulations. Moreover, as detailed below, all operators involved in a pending case will be required to submit a rate justification on the required FCC Forms within 30 days after the revised rules take effect on May 15, 1994. 145. We believe that there are generally five points in the rate-setting process that a local franchising authority and cable operator may be on May 15, 1994. The first is where the franchising authority has not certified to regulate basic rates by that date, or has certified but has not yet notified the operator that it is commencing basic rate regulation. In this case, the cable system will be required to file both an FCC Form 393 and new FCC Form 1200 30 days after the local authority notifies it that the authority is initiating rate regulation of the basic service tier. FCC Form 393 will be used to determine the operator's permitted rates from September 1, 1993 until May 15, 1994, and FCC Form 1200 will be used to determine its permitted rates after May 15, 1994. The franchising authority will then be expected to examine both filings within the timeframes established in Section 76.933. 146. Second, if, by May 15, 1994, a franchising authority has notified the cable operator that it has become certified, but the operator has not yet submitted the required FCC form (e.g., because the 30 day response period has not lapsed), we will require the cable operator to file both FCC Form 393 and new FCC Form 1200 within 30 days after May 15, 1994. We believe this limited deviation from Section 76.930 of our rules, which requires the cable operator to file its schedule of rates within 30 days of the date of a written request from the franchising authority, is justified because it will allow the operator to complete both forms simultaneously, with minimal disruption to the franchising authority. The franchising authority will be expected to examine both filings within the time periods established in Section 76.933. 147. Third, if, by May 15, 1994, a franchising authority has received the cable operator's filing for justifying rates under the old benchmark system but has not reached a final decision pursuant to Sections 76.933 or 76.936 of our rules, we expect the franchising authority to follow all existing timeframes with respect to that part of the proceeding. The cable operator in this situation will also be required to file new FCC Form 1200 with the franchising authority within 30 days of May 15, 1994. The franchising authority will then resolve the second portion of the proceeding, in which it will evaluate the cable operator's rates under the revised rules, according to the separate timeframes established in Section 76.933. 148. Fourth, if, by May 15, 1994, a franchising authority has reached a final decision about the lawfulness of an operator's basic rates under our initial rate rules, we will require the cable operator to file new FCC Form 1200 within thirty days of May 15, 1994. The franchising authority will then examine the form pursuant to the time frames established in Section 76.933. 149. Finally, if a franchising authority has reached a final decision on the cable operator's rates, and a rate increase request is pending as of May 15, 1994, the cable operator will be required to file new FCC Form 1200 within thirty days of May 15, 1994. The rate increase request will then be evaluated pursuant to the data submitted on the FCC Form 1200, rather than by any data that had already been submitted by the operator in support of its rate increase request. h. Pending Complaints Before the Commission 150. There are currently a number of complaints regarding rates for cable programming services tiers pending before the Commission. Unless we have issued a decision on a pending complaint before May 15, 1994, the operator about whom the complaint was made must file an FCC Form 1200 in addition to the FCC Form 393 it either has filed or must file. Such operators will be subject to refund liability calculated under our initial regulations for rates that were in effect from September 1, 1993 until May 15, 1994 (although obviously any refund liability will not start until the complaint was filed). Refund liability with respect to rates charged on and after May 15, 1994 will be calculated pursuant to the revised rules adopted in this Order. 151. Our reason for imposing this requirement is that most operators whose rates did not comply with our former rules also will be in violation of our new rules due to the increased competitive differential. Therefore, in most cases, a valid complaint made prior to the effective date of these new rules will remain valid. In addition, rates that may have been permissible under our former rules may be in violation of our new rules. Requiring operators against whom a complaint has been filed to submit an FCC Form 1200 thus will enable us to determine whether those operators are currently complying with our rules, regardless of their status under our initial rules. 152. To the extent there is a complaint regarding cable programming service tier rates pending before the Commission, we will continue to require the cable operator in question to file notice of any changes in rates with the Commission. This notice must be filed at least 30 days before such rates are proposed to be effective. This notice is necessary to allow the Commission to ensure that the cable service tier rate is not unreasonable. 5. Commission Authority to Adopt the Modified Ratemaking Approach 153. We conclude that the modified ratemaking approach we now adopt is consistent with our statutory authority under the Cable Act of 1992. Section 623(b)(1) of the Communications Act, 47 U.S.C. Section 543(b)(1), mandates that the Commission ensure that rates for the basic service tier are "reasonable." In addition, regulated upper tier rates may not be "unreasonable." Section 623(c)(1), 47 U.S.C. Section 543(c)(1). Nothing in the Cable Act of 1992 compels the use of a specific ratemaking model to ensure that rates are reasonable. Rather, Section 623(b)(1) of the Act, 47 U.S.C. Section 543(b)(1), specifies that the "goal" is "protecting subscribers" from higher rates than would be charged if all cable systems were subject to "effective competition." Sections 623(b)(2)(C) and 623(c)(2) of the Act, 47 U.S.C. Sections 543(b)(2)(C) and 543(c)(2), also set forth various factors the Commission must consider in establishing its ratemaking approach, but the statute leaves to the Commission the way in which these factors should be taken into account. 154. Indeed, Congress specifically rejected the approach mandated by the House bill, which would have required the Commission to establish a formula to set the maximum price for basic tier service. As enacted, Section 623(b)(2)(B) of the Act, 47 U.S.C. Section 543(b)(2)(B), instead provides that the Commission "may adopt formulas or other mechanisms and procedures." The legislative history explains: Rather than requiring the Commission to adopt a formula to set a maximum rate for basic cable service, the conferees agree to allow the Commission to adopt formulas or other mechanisms and procedures to carry out this purpose. The purpose of these changes is to give the Commission the authority to choose the best method of ensuring reasonable rates for the basic service tier and to encourage the Commission to simplify the regulatory process. Thus, the statute does not require us to craft a benchmark system of regulation, to calculate an estimated competitive differential, or to use any particular weighing methodology to calculate the competitive differential. 155. More generally, the courts have recognized that regulatory agencies are afforded wide latitude in discharging their ratemaking functions. Rather than insisting upon a single regulatory method for determining reasonable rates, courts evaluate whether the "end result" of a particular regulatory scheme results in rates that are within a "zone of reasonableness." Ratemaking under the Communications Act and similar statutes "is not an exact science." It involves both quantitative and qualitative judgments and predictions about the future. Thus, courts have stressed that "neither law nor economics has yet devised generally accepted standards for the evaluation of ratemaking orders." Moreover, as one court has noted in reviewing the exercise of our ratemaking power in the telephone context, ratemaking decisions "are appropriately treated as policy determinations in which the agency is acknowledged to have expertise." Thus, the courts recognize that regulatory agencies generally have broad discretion to choose methods and procedures in ratemaking determinations, provided the rates are within a "zone of reasonableness." By choosing to require that basic rates be "reasonable" and that upper tier rates not be "unreasonable," we believe that Congress has invoked this general body of law for application under the Cable Act. Thus, we have been instructed to consider the factors enumerated in the Cable Act and to use our expertise to achieve Congress's overall goal of ensuring "reasonable" rates for subscribers. 156. Against this legal backdrop, we have reconsidered our overall methodology for achieving the statutory goal of setting reasonable rates. While our initial regulatory methodology furthered the goal of the statute, we believe that methodology can be improved, for the reasons explained in detail herein. Upon further reflection, and based on a fuller record, we now believe that our revised methodology better achieves the overriding statutory goal of establishing reasonable rates. 157. In this regard, in response to NYNEX's petition for reconsideration, we have revisited and refined our approach to analyzing the competitive sample for purposes of estimating the competitive differential. Although we disagree with NYNEX that we can or should exclude the rates of low penetration systems from our competitive samples, we believe that our revised approach more appropriately considers the rates charged by all three categories of systems (low penetration systems, overbuilds, and municipals) that are deemed to be subject to effective competition under Section 623(l)(1) of the Act, 47 U.S.C. Section 543(1)(1). 158. In our view, our revised competitive differential of 17 percent is a more accurate reflection of the overall competitive differential, and based on a sounder methodology, than the figure of ten percent that we previously used. The difference between the two figures is primarily due to the fact that we previously averaged all of the data in our competitive sample, thus giving disproportionate weight to the low penetration systems which constituted more than half of the sample, while we have now taken a more qualitative approach, under which the data regarding the three groups are considered separately. An averaging approach is less appropriate, as a technical matter, than the qualitative approach we have now adopted because the statistical tests show that the three types of systems differ from each other. Where different groups are statistically different from each other, there is no good reason to consider them as one group. It makes more sense to consider them separately, since the statistical tests tell us that they are separate groups. Our refinement illustrates the sort of "pragmatic adjustment" that is within our discretion under the Cable Act of 1992 and general ratemaking principles. In this manner, we more fully effectuate the statutory goal of establishing reasonable rates. 159. While we have modified the way in which the three statutory classes of systems deemed to be subject to effective competition are taken into account in arriving at the competitive differential, we have continued to consider all three categories. Our goal, as before, is to set "reasonable" rates that are no higher than the rates of systems subject to "effective competition" as defined by Congress. We believe our modified competitive differential establishes more "reasonable" rates, that is, rates approximating what would be charged if cable systems faced effective competition. 160. Our refined approach is consistent with our previous determination that "cable systems with less than 30 percent penetration should continue to be included in the sample of systems subject to effective competition." While we have determined that we may not exclude from the competitive sample the rates of one of the category of systems that Congress has deemed to be subject to "effective competition," nothing in the Cable Act of 1992 requires that we estimate a competitive differential simply by averaging the per-channel rates charged by all of the systems included in our competitive sample and comparing that average to the average per-channel rate charged by the systems in our noncompetitive sample. As stated above, Congress decided not to enact the House proposal that would have required us to establish formulas, instead providing that we "may adopt formulas or other mechanisms and procedures." In addition, giving more weight to the data relating to overbuild systems -- systems that actually compete against one another to some extent -- is consistent with Congress's finding that "[w]ithout the presence of another multichannel video programming distributor, a cable system faces no local competition," and "[t]he result is undue market power for the cable operator as compared to that of consumers." 161. At the same time, it is important to note that the data relating to low penetration systems affected the calculation of the competitive differential. If instead of not differing from the noncompetitive systems in our sample, the rates charged by low penetration systems had been significantly lower, we might have selected a higher competitive differential. For example, if the rates of low penetration systems in areas with many broadcast television stations had been 25 percent lower than the rates charged by noncompetitive systems, that would have suggested that the availability of over-the-air broadcast television signals limits the market power of cable systems in some circumstances. Since it is hard to see why overbuilds would have more market power than low penetration systems, a 25 percent differential would have suggested that the differential observed in our sample between overbuilds and noncompetitive systems was depressed by collusion to a greater extent than we have estimated. Accordingly, we would have revised that estimate or placed more reliance on the data involving municipal systems if the data relating to low penetration systems had been different. With respect to our authority under the statute, the point is that it is clear that the rates charged by low penetration systems were "take[n] into account," since the rates that were observed for low penetration systems affected the calculation of the competitive differential. 162. Similarly, the other factors that we have been directed to "take into account" or "consider" will affect the rates charged by regulated operators. Those factors primarily involve the costs and revenues of cable companies. Because we recognize that application of the competitive differential would not result in a reasonable rate for every cable system, but would instead set the rates of some systems below the amount that a cable operator without market power would charge, we are today adopting revised "cost-of-service" regulations that will permit cable operators to choose not to apply the competitive differential, but instead to have their rates set according to procedures analogous to those used to set the rates of public utilities. The optional "cost-of-service" rules that we announce today are based largely on the costs and revenues of cable companies. The existence of the cost- of-service "safety valve" affects our determination of the competitive differential by allowing us to estimate it most accurately, secure in the knowledge that those operators for whom our competitive differential is inaccurate may choose not to use it. 163. Moreover, the factors involving the costs and revenues of cable companies are taken into account more directly in our calculation of the competitive differential. In determining whether a rate is "reasonable," we have focused on whether a cable operator without market power would charge such a rate. In order to stay in business over the long run, a cable operator without market power would need to charge rates that would cover, for example, its "direct costs . . . of obtaining, transmitting, and otherwise providing signals," and "the capital and operating costs of the cable system," which are among the factors we have been directed to consider in setting rates. Likewise, it is reasonable to assume that a cable operator lacking market power would charge rates that reflect, for example, "the revenues (if any) received by a cable operator from advertising." The reasonable rate that is determined by applying the competitive differential should allow cable operators to cover their costs and obtain a reasonable profit. At the same time, the cost-of-service option is available for cable operators that would not be able to cover their costs after applying the competitive differential. 164. In certain respects, the approach we now adopt is analogous to the judicially approved manner in which we prescribe a rate of return for telephone companies. In the telephone context, we select a prescribed rate of return from within a broad "zone of reasonableness" that is bounded generally on the upper end by rates that would be unreasonably high from the perspective of consumers and on the lower end by rates that would not sufficiently protect the interests of investors in the regulated enterprise. In selecting the prescribed rate of return within this broad range of permissible rates, we consider numerous factors that go into the ratemaking decision, according greater or lesser weight to individual factors on the basis of the record and in the exercise of our judgment and expertise. 165. That is essentially what we have done in selecting the revised competitive differential. We have examined the rates charged by the three types of systems that Congress has determined to be subject to effective competition and determined that they charge between one percent and 37 percent less than noncompetitive systems. We have narrowed that broad range by focusing on the overbuild sample, since we have determined that the rates charged by overbuilds best approximate the "reasonable" rates that would be charged by systems lacking market power. We have adjusted the differential between overbuilds and noncompetitive systems in light of the data relating to low penetration systems and municipals, the lack of full head- to-head competition between overbuilds, and the possibility of collusion between operators in an overbuild situation. Those adjustments substantially restricted the zone from which the competitive differential was selected. 166. Our current approach is consistent with our prior legal analysis, since we did not say that all of the specifics of our prior approach were mandated by the statute. Nor could we plausibly have contended that the statute required us to calculate a competitive differential by relying on a simple averaging of the rates of all systems subject to "effective competition" as defined in the Act. The only possible basis for such a conclusion would be the sentence in Section 623(b)(1) of the Communications Act, 47 U.S.C. Section 543(b)(1), which provides that our "regulations shall be designed to achieve the goal of protecting subscribers of any cable system that is not subject to effective competition from rates for the basic service tier that exceed the rates that would be charged for the basic service tier if such cable system were subject to effective competition." But it would not be a natural reading of that sentence, which appears to emphasize the goal of "protecting subscribers," to conclude that it requires us to calculate a competitive differential following the precise methodology that we followed previously. Besides being an unnatural reading of that sentence, such an interpretation would be inappropriate for a number of other reasons. First, as noted above, Congress made clear in Section 623(b)(2)(B), 47 U.S.C. 543(b)(2)(B), that it was not mandating the use of formulas at all. In addition, such an interpretation would negate Congress's general instruction directing us to establish "reasonable" rates. That is, there would be no point in instructing us to set "reasonable" rates if we were compelled to set rates by calculating the competitive differential exactly as we did before. Furthermore, in Section 623(b)(2)(C)(i), 47 U.S.C. Section 543(b)(2)(C)(i), Congress instructed us to "take into account . . . the rates for cable systems, if any, that are subject to effective competition." There would be no need to instruct us to take those rates "into account" if another provision compelled us to use them to set rates in a specific manner. In addition, the "if any" language in Section 623(b)(2)(C)(i) suggests that Congress was not sure that any cable systems were subject to "effective competition," and a Congress that was unsure of that fact would not have compelled us to follow the methodology that we followed previously, which is entirely dependent on the existence of such systems. Finally, it is not clear why Congress would have instructed us to take various other factors "into account" if we were obliged to follow a formula set out in another subsection of the statute. 167. In addition to adjusting the competitive differential, the other key change made with respect to rate calculations in this Order is our decision to apply the competitive differential to most cable systems although immediate rate reductions will not be required for operators with relatively low rates or for small operators while we study the prices and costs these operators experience. Nothing in the statute suggests that we were required to use a benchmark approach. Nor is our decision to defer rate reductions for some operators, pending completion of our cost study, unreasonable. To the contrary, as we have explained, nothing in the current record suggests that the competitive differential should not be applied to all regulated operators, and economic theory suggests that cable operators with market power will exercise it. The only factor favoring an approach that limits immediate reductions is our concern that cable operators with relatively low rates may not be exercising market power to the same degree as those with higher rates. Absent cost data, we cannot determine whether that concern warrants a revision of the competitive differential for cable operators with relatively low rates. In those circumstances, the competitive differential ought to be applied to all regulated operators unless cost data convinces us otherwise. 168. Nor is our decision to delay application of the competitive differential to small operators, pending a cost study, unreasonable or contrary to the statute. As we have explained, we lack reliable data with respect to the financial situation of small operators as a class. In addition, small operators are more vulnerable than larger operators, and therefore a small operator may be harmed more seriously by application of the competitive differential if the competitive differential is not appropriate. Of course, small operators may invoke cost of service procedures -- indeed, a streamlined cost of service procedure is available for small operators -- but it would be administratively burdensome if regulators were flooded with cost of service applications from small operators. In those circumstances, it is prudent to give small operators a further opportunity to present cost data showing that they do not, as a rule, charge unreasonable rates, and to stay application of the competitive differential for small operators until that study is completed. 6. The Price Cap Governing Cable Service Rates a. Calculation of External Costs 169. In our April 1993 Rate Order, we determined that rates for regulated cable services would be governed by a price cap once initial regulated rates were set. We also determined that the price cap should be expressed as a price per channel. 170. In the Rate Order, we determined that capped rates may be adjusted for inflation. The Commission selected an annual adjustment measured by the gross national product fixed weight price index (GNP-PI). We determined that cable operators could make such inflation adjustments annually, but stated that cable operators could adjust for inflation for the part of the year between the initial date of regulation and the beginning of the next year. 171. The Rate Order also provided that cable operators may pass through to subscribers increases in certain categories of external costs. Those external costs included the additional or new retransmission consent fees incurred after October 6, 1994, other programming cost increases (with some exceptions for program purchases from affiliates), taxes, franchise fees, and the costs of other franchise requirements including the costs of any public, educational, or governmental access programming required by the franchising authority. We concluded that external cost recovery (except for franchise fees to which the annual inflation adjustment did not apply) would be permitted only to the extent that the increases exceed the rate of inflation. We also provided that, with the noted exception for retransmission consent fees, cable operators could pass through to subscribers any changes in external costs that accrued after the date on which the tier at issue became subject to regulation or 180 days after the effective date of our initial regulations (e.g., February 28, 1994), whichever occurred first. 172. We subsequently decided that operators could file rate increases no more than quarterly on account of external cost increases. We also decided that an operator must reduce its permitted rates to reflect any decreases it experiences in external costs. Such decreases must be reflected in any filings the operator makes for inflation or increases in external costs and, in any event, all decreases must be reflected in the operator's rates within one year from when they occurred. 173. Thus, under our initial rules, operators may adjust their regulated rates annually by inflation and up to quarterly by the net change in external costs. As indicated above, any change in external costs must also be measured against inflation and adjusted for the corrected inflation rate. In order to simplify the calculations to be used for making these rate adjustments, however, we reconsider on our own motion our rules in this area. 174. Specifically, we have decided to separate the inflation adjustment from the external cost adjustment. Under the new approach, an operator seeking to adjust capped rates to reflect changes in its external costs will determine the actual level of its external costs. This figure will then be removed from the total charge for the affected service tier. The "residual" will be adjusted for inflation on an annual basis, but no earlier than September 30 of each year, when the final GNP-PI through June 30 of each year is released, and no later than December 31 of each year. By contrast, the external cost component that does not include the "residual" may be adjusted quarterly for net changes in external costs. 175. Because the residual rate that is adjusted for inflation does not reflect the operator's external costs, there is no possibility of double recovery of external cost increases that concerned us in the Rate Order. Therefore, it will no longer be necessary to compare changes in external costs to inflation, as under our initial approach, or to later adjust the external cost increases based on the annual inflation rate. The new approach will produce the same rates as the requirements specified previously in the Rate Order, but should be simpler to apply. We therefore adopt it as the method for making adjustments to capped rates for inflation and changes in external costs. 176. We will retain the requirement that, in the absence of a showing that a rate increase is necessary to avoid confiscation, operators may file rate increases no more frequently than quarterly to reflect increases in external costs, under the requirements specified in the First Recon. Order. However, to simplify the filing procedures significantly, we are requiring all systems to use calendar year quarters, rather than quarters that begin on the date the tier at issue became subject to regulation. We also will allow operators to accrue external costs for any program service tier from the date on which the first of the operator's tiers became subject to regulation (or February 28, 1994, whichever was earlier). This will allow operators to keep track of their external costs from a single date rather than from multiple dates that may or may not be close in time to each other. And, we are clarifying that operators may file for a rate increase on account of changes in external costs as soon as the information necessary to make the change is available. We believe this process best serves the interests of both subscribers and operators, without creating undue administrative burdens. 177. FCC Form 1210 and associated instructions set forth the specific steps for making these calculations. b. Copyright Fees 178. The Commission has not addressed whether copyright fees incurred by the carriage of distant broadcast signals should be accorded external cost treatment. Several petitioners request that we treat such fees as external costs, and that we allow systems to recover copyright fees from September 30, 1992, forward, separate and apart from the operators' permitted rates. In support of their request, these petitioners assert: (1) that copyright fees were not reflected in the Competitive Survey data on which we based the rate formula, because many operators separately itemize these costs on subscriber bills; (2) that copyright fees should be treated like taxes because they amount to a government-imposed tax on the transaction between operators and subscribers; and (3) that if copyright fees are not given external cost treatment, many operators will be forced either to drop distant broadcast signals or submit cost-of-service showings. 179. To the extent that petitioners are asking that we revise the rate calculation process by subtracting copyright fees from the rate survey data and then permitting a system's specific copyright fees to be added back to calculate the allowable rate for a specific system, we do not believe this change is either practical or justified. Although copyright fees may be separately itemized by some systems, there is no evidence in the record that operators did not include these itemized fees when providing information about their rates in response to our rate survey. Thus, there is no reason to assume that our survey results do not already reflect copyright fees, or that the requested change would make our survey results more accurate. Moreover, we previously determined that copyright fees do not constitute taxes on the transaction between the operator and the subscriber because (1) they are imposed only in the sense that all legal obligations are imposed by government; and (2) these fees are merely consensual arrangements relating to the consideration to be paid in exchange for carriage of programming. We thus are unpersuaded by petitioners' argument that copyright fees are "taxes" that should therefore be given external cost treatment. 180. However, to the extent that petitioners' argument is that increases in compulsory copyright fees incurred by carrying distant broadcast signals should be treated in a fashion parallel to increases in the contractual costs for nonbroadcast programming, we believe it has merit. Section 76.922(d)(2) of our regulations, subject to specified limitations, permits subscriber rates to be adjusted to take into account changes in certain "external costs," including "programming costs." Copyright fee increases, whether they result from the addition of new broadcast signals to a tier, adjustments to the fee levels by the arbitration panels under the aegis of the Copyright Office, or from adjustments in tier structures, appear to fit logically within the programming costs category. c. Pole Attachment Fees 181. The Commission has not addressed whether pole attachment fees should be accorded external cost treatment. Some cable operators argue that costs associated with pole attachment fees should be treated as external costs because (1) such fees are largely beyond their control since operators typically must utilize existing poles owned by utilities and have no choice but to pay any utility-imposed increases in pole-attachment fees, and (2) pole attachment fees historically have increased at a rate faster than inflation. 182. Although pole attachment fees are to some extent beyond the control of system operators, we are not persuaded that they are sufficiently unique among the numerous categories of costs that make up the expense accounts of system operators to warrant external treatment. Unlike increases in franchise fees or taxes, pole attachment fees are not imposed by the government nor are they, like programming expenses, an area with respect to which the legislative history of the 1992 Cable Act expresses explicit concern. In addition, some pole attachment fees are regulated under the 1978 Pole Attachment Act, 47 U.S.C. ÀÀ 224, which should provide operators some recourse against unreasonable pole attachment fee increases. We accordingly will not permit operators to treat pole attachment fees as external costs. 7. Other Rate Issues 183. Other rate issues raised in petitions for reconsideration are (1) whether cable operators should be allowed to charge commercial customers different rates than residential customers, (2) whether systems located in Alaska and Hawaii should be exempted from benchmark regulation, and (3) whether we should adopt a subscriber line charge that would be paid by consumers purchasing only basic service. a. Commercial Rates 184. Cablevision and NCTA urge the Commission to clarify that cable operators are not required to charge the same rates to residential and commercial customers. They contend that the ability to charge commercial rates will facilitate the provision of service to commercial subscribers, such as sports bars and restaurants, and that this benefits viewers who are customers of these commercial establishments. Cablevision also states that the 1992 Cable Act does not specifically require that the same rates be applied to commercial and residential customers. It argues that certain isolated references in the legislative history to "homes" and "households" indicate an intent that residential service be regulated differently than commercial service. Cablevision requests that cable operators be allowed to charge different rates to commercial subscribers, and that the determination of whether such rates are reasonable or unreasonable be made in the context of the commercial use of such cable service. 185. We are not persuaded that the Commission should establish provisions authorizing special, presumably higher, rates for regulated cable services provided to commercial establishments. While it is possible that our overall standards for reasonable rates could, depending on the specific implementation, provide for special, higher rates for commercial establishments, neither the Cable Act nor its legislative history evinces an intent that the Commission should generally do so. Moreover, petitioners have not suggested how such rates would be determined. Nonetheless, it is possible that higher rates for commercial establishments could play a role in assuring that rates for subscribers are reasonable if the higher commercial earnings were offset by savings to subscribers. Accordingly, while we will not adopt regulations permitting special commercial rates at this time, we will consider, on a case-by-case basis, specific proposals that cable operators may want to make that would produce savings for consumers. In addition, we are further exploring this issue in our Further Notice of Proposed Rulemaking. b. Rate Relief for Alaska and Hawaii 186. The Medium-Sized Operators Group requests that the Commission establish special rates for cable systems located in the Alaska and Hawaii. It claims that system operating costs are significantly higher for cable systems in Alaska. Alaska Cablevision urges an exemption from rate regulation for Alaskan cable systems, particularly small systems serving rural areas. 187. As discussed above, we have determined that the rates of cable systems not subject to effective competition as defined in the 1992 Cable Act presumptively reflect their market power. With certain transitional exceptions, we have required that all regulated operators reduce rates by the competitive differential in order to avoid refund liability. Petitioners have failed to present any evidence showing that rates for cable service provided by operators subject to regulation in Alaska and Hawaii do not also reflect their market power. Thus, we see no reason why operators in those states should not be subject to the same competitive differential reduction as operators in other states, or why consumers in those states should not see benefits comparable to those that will be experienced by consumers in other states. 188. Moreover, the rates charged by operators in an unregulated marketplace in Alaska and Hawaii presumptively permitted recovery of any higher costs of providing service in those states. Application of the competitive differential will reduce cable rates in those states by the same percentage as in other states. Thus, operators in Alaska and Hawaii will be able to maintain rates that reflect any relatively higher costs of providing service in those states, although reduced by the competitive differential. Furthermore, the petitioners have failed to present cost information showing that, or to what extent, the costs of providing cable service in Alaska and Hawaii are higher than in the other forty-eight states. Therefore, we are unable in any event under the present record to fashion adjustments to rates to address allegedly higher costs of providing cable service in Alaska and Hawaii. We thus reject petitioners' requests on this issue. Of course, cable operators in Alaska and Hawaii with unusually high costs may invoke the cost-of-service rules. In addition, small and low priced cable operators in Alaska and Hawaii will be eligible for transition treatment on the same terms as other cable operators in the rest of the country. c. Basic Tier Access Charge 189. Under our methodology for setting initial regulated rates, operators develop per channel rates that are averaged across all regulated tiers, i.e., the development of rates is tier neutral. The Medium-Sized Operators Group contends that tier neutrality severely limits an operator's ability to recover the costs associated with providing the basic service tier. Petitioner claims, without support, that the cost per subscriber to provide the capital and to maintain the cable infrastructure is approximately $20 per month, excluding programming costs. Petitioner thus proposes that we adopt a "subscriber line charge" that would be paid by subscribers who purchase only the basic tier. It contends that such an approach would be consistent with policies adopted by the Commission in connection with common carriers, citing MTS & WATS Market Structure: Third Report and Order, 93 FCC 2d 241 (1983). 190. Petitioner has failed to address specifically why our requirements for setting rates for the basic and upper tiers do not adequately permit recovery of all costs of providing cable service properly allocated to regulated tiers. Petitioner has also failed to demonstrate that our approach does not permit the recovery of all costs that could be assigned to a tier under any reasonable allocation method. In addition, we note that the federally mandated subscriber line charge to which petitioner refers was established to properly allocate to telephone subscribers a portion of nontraffic sensitive fixed costs associated with local loop plant. Petitioner has not demonstrated that there are similar costs associated with the provision of basic cable service. Accordingly, we reject petitioner's assertion that operators need a special surcharge on basic service. C. "A La Carte" Packages 191. Under the 1992 Cable Act, video programming offered on a per channel or per program ("a la carte") basis is not subject to rate regulation. In the April 1993 Rate Order, we held that we would not regulate collective offerings of otherwise exempt per channel or per program services so long as: (1) the price for the combined package does not exceed the sum of the individual charges for each component of service, and (2) the cable operator continues to provide the component parts of the package to subscribers separately. We stated that the second condition would be met only when the per channel offering provides subscribers with a realistic service choice. We also stated that we would retain jurisdiction to review individual offerings of "a la carte" channels to determine whether the attempted offering constituted an evasion of rate regulation. 192. In light of the limited type of collective offerings of per channel services that were available at the time we adopted the Rate Order, we believed it was in the public interest to allow operators to sell collective offerings of "a la carte" services without subjecting the offering to rate regulation. In particular, we believed that market forces would likely ensure that the rates for these offerings would be reasonable. Allowing nonregulated treatment of collective offerings of "a la carte" channels could also serve consumers' interests by making it possible for them to purchase packages of unregulated channels at a lower price than if channels were purchased individually. 193. However, since the adoption of the Rate Order, a number of operators have restructured service offerings so that channels that could have been subject to regulation have been removed from a regulated tier and are now offered on an "a la carte" basis as well as on a package basis. Since the rates of the collective offerings of the "a la carte" channels are unregulated, operators may raise their overall rates for the same service by removing channels from regulated tiers and offering them on a package and an "a la carte" basis. We are concerned that this practice may not be consistent with the purposes of the 1992 Cable Act. In addition, we have received numerous complaints from local franchising authorities and subscribers concerning the terms and conditions of "a la carte" offerings of channels. We are concerned that some of these offerings may not comply with our requirement that subscribers must have a realistic option to purchase channels that are not subject to regulation on an "a la carte" basis. Some of the repackagings of channels may also constitute prohibited evasions of rate regulation. 194. As stated in the Rate Order, providing for nonregulated treatment of collective offerings of "a la carte" channels affords operators an opportunity to enhance consumer choice by making programming more affordable and more widely available. At the same time, however, permitting nonregulated treatment can provide an opportunity for operators to engage in evasion of rate regulation. On reconsideration, therefore, we continue to believe that the public interest will be served by generally permitting nonregulated treatment of collective offerings of "a la carte" channels if the offering enhances consumer choice and does not constitute an evasion of rate regulation. We believe that these objectives will be achieved if operators comply with the safeguards of our initial rules. Thus, collective offerings of otherwise exempt per channel or per program services will continue to be unregulated as long as: (1) the price for the combined package does not exceed the sum of the individual charges for each component of service, and (2) the cable operator continues to provide the component parts of the package to subscribers separately. This latter safeguard will be met if the "a la carte" offering constitutes a realistic service choice. 195. However, in order to address our concerns that some offerings established by operators in response to rate regulation may not be consistent with our goals and the 1992 Cable Act, and the fact that other offerings that could raise similar concerns could be initiated in the future, we are here providing interpretive guidelines for determining whether an operator's collective offering of "a la carte" channels should be accorded regulated or unregulated treatment. These guidelines will enable operators to better determine what collective offerings of "a la carte" channels will be considered an evasion of rate regulation and/or a realistic service offering, and will help local authorities and the Commission to assess expeditiously the appropriate regulatory status of individual offerings. In evaluating offerings in individual cases, we will consider whether consumers are being offered a greater variety of programming choices and options and whether the price for those choices is generally increasing or decreasing from previous levels. 196. We have identified several factors that local authorities and the Commission should consider in assessing in an individual case whether an "a la carte" package enhances consumer choice and does not constitute an evasion of rate regulation. Several of these factors, if present, would suggest that the rates for the offering should be unregulated. These are: (1) the operator had offered (or begun to explore offering) "a la carte" packages consisting of non- premium channels prior to rate regulation; (2) the operator has conducted market research that suggests introducing an "a la carte" package" would be profitable, other than as a means of evading rate regulation; (3) the subscriber is free to select which channels will be included in the package; (4) subscribers are given notice that fully discloses their options, as well as fully discloses the total price (including related equipment charges) associated with exercising any of these options; and (5) an insignificant percentage or number of channels in the package has been removed from regulated tiers. On the other hand, the following factors would weigh against allowing unregulated treatment of collective offerings of "a la carte" channels: (1) the introduction of the "a la carte" package results in avoiding rate reductions that otherwise would have been required under the Commission's rules; (2) a significant percentage or number of channels in the package were removed from regulated tiers; (3) the package price is so deeply discounted when compared to the price of an individual channel or the sum of the prices of the individual channels that it does not constitute a realistic set of service choice because subscribers will not have any realistic options other than subscribing to the package; (4) the channels taken from regulated tiers have not traditionally been marketed "a la carte"; (5) an entire regulated tier has been eliminated and turned into an "a la carte" package; (6) the subscriber must pay a significant equipment charge to purchase an individual channel in the package; (7) the subscriber must pay a "downgrade charge" (an additional charge) to purchase an individual channel in the package; (8) the "a la carte" package includes channels that were removed from lower tiers of channels, so that subscribers to those lower tiers are required to buy one or more intermediate tiers in order to receive the same channels; (9) subscribers are automatically subscribed to the "a la carte" package through, for example, such means as negative option billing; and (10) the affected programmers object to the restructuring of their services into "a la carte" packages. No single factor will necessarily be dispositive in any case. Rather, we will assess the totality of the circumstances, analyze whether one or more of the foregoing factors is present, and determine whether the offering intentionally, or in effect, constitutes an evasion of rate regulation. 197. Under our requirements for setting initial regulated rates, the rate for a regulated program service tier will be determined in part by the system's total revenues for all regulated tiers and the total number of regulated channels it offers. Similarly, our methodology for adjusting capped rates for adding or deleting channels from a regulated tier is determined in part by the total number of regulated channels. Thus, local authorities will need to determine the total number of regulated channels offered by the operator in order to properly set rates for the basic service tier. This will require a determination of whether any collective offerings of "a la carte" channels should be considered a regulated tier. Accordingly, we will permit local authorities to make initial determinations as to whether a collective offering of "a la carte" channels should be considered a regulated tier, even if the collective offering would be a cable programming service tier if it were regulated. 198. Under the requirements that we are adopting, local authorities may, at their option, make an initial decision addressing only the regulatory status of any "a la carte" package at issue. The franchising authority must make this initial decision within the 30 day period for reviewing basic cable rates and equipment costs, or within the first 60 days of an extended 120 day period (if the franchising authority has requested an additional 90 days). The franchising authority shall provide public notice of its initial decision within seven days pursuant to local procedural rules for public notice. Operators or consumers may make an interlocutory appeal of this initial decision to the Commission within 14 days of the initial decision. (Within 14 days of the initial ruling, an operator shall provide notice to the franchising authority whether it will, or will not, make such an appeal.) The Commission will rule expeditiously on these appeals, and the local authority may then proceed with its local rate case in light of the Commission's decision on the interlocutory appeal. 199. Alternatively, local authorities may make any necessary "a la carte" determination as part of their final decision setting rates for the basic service tier. That decision may then be appealed to the Commission as provided under current rules concerning appeals of local decisions to the Commission. In any appeal of a local decision, the Commission will defer to the local authority's findings of fact if there is a reasonable basis for the local findings. The Commission will then apply FCC rules and precedent to those facts to determine the appropriate regulatory status of the tier in question. Local authorities may also request that the Commission make the initial "a la carte" decision by means of a petition for declaratory ruling. These provisions for local determination of "a la carte" issues will facilitate local authorities setting rates for the basic service tier while providing for Commission oversight of local decisions that could affect the regulatory status of cable programming services tiers. 200. We will monitor our treatment of collective offerings of "a la carte" channels. If it appears that they are not adequately fulfilling the purposes of the 1992 Cable Act, we will promptly revisit them. D. Small Systems 1. Small System Administrative Relief a. Background 201. The Cable Act of 1992 requires the Commission to develop and prescribe cable rate regulations designed "to reduce the administrative burdens and cost of compliance for cable systems with 1,000 or fewer subscribers." In implementing this section, we are guided both by the explicit statutory mandate and by the view that regulatory requirements should generally be no more burdensome than necessary to achieve full compliance. We are particularly concerned about the administrative burdens imposed on small systems, whose more limited revenue base and smaller clerical staff may make them less able to absorb administrative costs. 202. We took a first step at providing administrative relief for small systems when we adopted our current rate regulations. Specifically, in the Rate Order, we authorized franchising authorities to permit small systems to certify that their rates for basic service and equipment are reasonable under our rate standards, and permitted (and encouraged) franchising authorities regulating the same small system to file joint certifications. We subsequently decided to take a more extensive look at the issue of regulatory relief for small systems, and at the same time stayed the effectiveness of rate regulation for small systems pending our review of the regulatory requirements applicable to them. We sought comment to supplement the record concerning possible additional measures to reduce administrative burdens on small systems. In addition, we asked whether we should differentiate between small systems owned by small multiple system operators ("MSOs") and those owned by larger MSOs when crafting administrative relief. b. Comments 203. In response to the Rate Order, many petitioners sought reconsideration of our initial effort to grant small systems some form of administrative relief. These petitioners contend that we did not take sufficient steps to ease the burdens of complying with rate regulation for small systems, as we had been directed to do by Congress. To support their claims, petitioners assert that cable rate regulation burdens fall disproportionately on small systems, which are least able to absorb the additional regulatory costs. They claim that small systems do not have the expertise, nor are they able to afford to hire experts, to prepare the analysis and forms necessary under rate regulation. They also claim that per subscriber costs are higher for small systems, which usually serve less densely populated rural areas and have smaller subscriber bases over which to spread costs. Moreover, they claim that small system operators must deal with more franchise authorities and face higher programming costs. Finally, they contend that revenue sources are limited for small systems. They attribute this financial limitation to the fact that small systems gain a smaller percentage of their revenues from unregulated programming services, such as premium channels and pay-per-view events. They contend that small systems generally do not have the technical ability to provide pay- per-view programming, and assert that advertising revenue is very limited as an alternative revenue source. 204. Most commenters responding to the Stay Order/Further Notice advocate treating small systems affiliated with multiple system operators ("MSOs") the same as independent small systems. These commenters contend that the 1992 Cable Act does not permit different treatment of MSOs and requires that all small systems receive relief from regulatory burdens. Moreover, commenters cite floor debate in the House where a proposal to increase the size of small systems from 500 to 1,000 subscribers was opposed because it would not exclude MSOs from its coverage, but was nevertheless adopted. These commenters further contend that all but the very largest MSOs lack large corporate administrative staffs, centralized data processing, billing, management, engineering and construction support, and volume discounts on programming and equipment. 205. Some commenters claim that regulatory burdens are particularly high on small MSOs, which may have dozens or hundreds of small systems, each subject to separate rate calculations. Moreover, they point out that the regulatory burden of preparing FCC Form 393 and/or a cost-of-service showing is incurred at the franchise level. Thus, they argue, small MSOs which have many small systems serving various franchise areas face regulatory costs equal to, or greater than, independent small systems. c. Discussion 206. Our rate regulations require systems electing the benchmark approach to setting rates to reduce their September 30, 1992 rates by the 17 percent competitive differential and then adjust the resulting rate forward for inflation, external costs and changes in the number of regulated channels offered by the system. Systems must also establish unbundled equipment charges based on actual costs. Completing our required benchmark forms can impose burdens on small operators, especially the various calculations required to unbundle equipment charges from program charges and set the former at actual cost. The record suggests that it is these equipment computations which impose the most significant burden in the rate-setting process. 207. Our rate requirements are designed to achieve rates for regulated cable service and equipment that will best fulfill the purposes of the Cable Act of 1992. We believe that subscribers of small systems should be able to obtain the benefits of these regulatory requirements to the same extent as subscribers of larger systems. At the same time, it may be more difficult for small systems to comply with these requirements in the same time frame as larger operators. Accordingly, on reconsideration, we are adopting rules in addition to those established in the Rate Order that will provide administrative relief for small systems but that will also, as described below, achieve substantial compliance with rate regulation requirements. (i) Streamlined Rate Reductions 208. As noted, under our revised benchmark rules, regulated cable systems must establish rates based on September 30, 1992 rates, as reduced by the competitive differential and then adjusted forward by inflation, changes in the number of regulated channels offered, and external costs. Generally speaking, when they become regulated, small systems are subject to these requirements unless they are eligible for transition relief. Such systems must also establish unbundled equipment charges based on actual cost. 209. In order to permit small systems to implement competitive rate reductions that will benefit subscribers while reducing administrative burdens, we will permit eligible small systems, as defined below, to reduce their rates under a streamlined approach instead of using the benchmark methodology set forth in FCC Form 1200. Specifically, small systems owned by eligible operators may elect to make rate reductions by reducing each billed item of regulated cable service by the competitive differential in lieu of following our more comprehensive rate regulation rules. Thus, operators may reduce the billed charge for each tier of regulated service by the competitive differential. The reduction will be from charges in effect as of March 31, 1994. The amount of the reduction will be 14 percent, which approximates the amount of the competitive rate reduction that would be applicable to many small systems that have not been regulated and that therefore did not implement any rate reductions under our initial rate regulations. 210. Similarly, under the requirements we are establishing today, small systems electing to reduce rates in this manner must apply the 14 percent reduction to each regulated equipment charge appearing on subscribers' bills. Thus, under this approach, eligible small systems are not required to unbundle equipment and installation charges from their programming service charges, or to set equipment and installation charges at actual cost. 211. We believe that these rate reductions will be easy to implement in that they merely involve application of a percentage reduction to each billable charge rather than the more extensive calculations set forth in FCC Form 1200. At the same time, the reductions will ensure that the subscribers of small systems pay reasonable regulated rates. 212. However, as stated, we believe that small systems ultimately can and should establish regulated rates on the same basis as other operators. Thus, this streamlined alternative to implementing rate reductions will be a temporary approach to setting rates prior to full compliance. In our Cost Proceeding, we have decided to examine whether we should establish possible average cost schedules for the provision of equipment, as well as average cost schedules generally for provision of regulated program service. These schedules could be used by small systems to set rates instead of requiring them to identify and evaluate their own costs or make the calculations required under our benchmark rules. Thus, these schedules could afford significant administrative relief for small systems as well as all systems generally. 213. We believe that when average cost schedules for equipment are developed, small systems that have elected to make streamlined rate reductions should be required to develop rates based on September 30, 1992 rates with specified adjustments as required of cable systems generally. This opportunity to implement streamlined rate reductions, subject to full compliance later, will give eligible systems a sufficient opportunity to prepare for compliance while affording consumers significant benefits in the interim. Thus, we will require small systems that have elected to make streamlined rate reductions to establish rates based on full compliance with our rules at the time we establish average cost schedules for equipment. 214. If a small system elects streamlined rate reductions, the permitted rate for a tier will be the rate for the tier in effect on March 31, 1994 minus 14 percent. The rate will then be capped at that level and the price cap requirements applicable to cable systems generally will govern rates for the tier. Thus, capped rates may be adjusted annually for inflation, quarterly for external costs, and quarterly for additions and deletions of channels. Small systems must use FCC Form 1211 when justifying such rate changes to local or federal regulators. 215. Our provisions for streamlined rate reductions are intended to provide administrative relief for small systems because small systems are less likely to have the resources to comply with our rate reductions in a timely fashion. On the other hand, some small systems as defined in the 1992 Cable Act may be affiliated with larger companies with the resources to fully comply with our rate requirements. Accordingly, we are establishing standards of eligibility for streamlined rate reductions designed to limit their availability to larger companies. 216. Streamlined rate reductions will only be available to independent small systems (i.e., those that are not owned by or affiliated with other cable systems) and to small systems owned by those MSOs that have 250,000 or fewer total subscribers, own only systems with less than 10,000 subscribers each, and have an average system size of 1,000 or fewer subscribers. The 250,000 subscriber cap ensures that larger MSOs do not benefit from a form of administrative relief that they may not need because they are likely to have the resources necessary to expeditiously establish compliance with rate regulation requirements. The 10,000 and 1,000 subscriber limits help tailor streamlined rate reductions to small systems that are owned by those operators that are most likely to own numerous small systems. The record strongly suggests that it is these operators that face the highest administrative costs of complying with our benchmark regulations because they own so many systems and all of their systems serve fairly small subscriber bases (i.e., none is larger than 10,000 subscribers). Streamlined rate reductions will not be available to any system that has already restructured its rates in an effort to comply with our rules, since such a system has demonstrated that it does not need the administrative relief that the streamlined rates reduction process is intended to provide. 217. Small systems electing to implement streamlined rate reductions must provide written notice to that effect to their subscribers, as well as to the local franchising authority with respect to the basic service tier and the Commission with respect to a cable programming service tier. This notice must be provided within 30 days after the small system becomes subject to regulation. The small system must then implement the streamlined rate reductions within 30 days after the notification has been provided. (ii) Company-Wide Averaging of Equipment 218. Under our benchmark rules, operators are required to aggregate expenses and revenues, including equipment and installation costs, at the franchise, system, regional or company level in accordance with the operator's practices as of April 3, 1993. In order to reduce administrative burdens associated with setting unbundled rates for equipment based on actual costs, we are permitting operators of small systems to average the equipment costs of its small systems at any level, or combination of levels, regardless of the operator's practices as of April 3, 1993, subject to safeguards designed to protect subscribers from unusual rate changes. Under this approach, a cable operator of any size may average the equipment costs of all its small systems, or only some of them, for purposes of developing unbundled equipment charges that are set at actual cost. This process will permit operators to select a level of averaging that enables them to develop equipment charges in accordance with our rules, but also involves the least administrative burden. Specifically, it may be less burdensome for some operators to average the equipment costs for their small systems at a level different from the level at which they generally averaged their costs on April 3, 1993. This flexibility therefore should reduce the overall effort and expense involved in separating equipment costs from programming costs when calculating individual small system rates. 219. We recognize, however, that setting equipment charges at a different level of cost averaging than the operator was employing on April 3, 1993 could involve rate changes both for equipment and programming service charges, since permitted rates for equipment and programming service charges are based on aggregate programming service and equipment charges as of September 30, 1992. In order to prevent sudden rate changes that could harm subscribers, we are establishing several safeguards that operators of small systems must follow when developing average equipment costs for those systems. First, the flexibility in averaging equipment costs will apply only to the operator's small systems, rather than the larger systems it owns. Second, it will only be permitted for equipment, as opposed to installation charges. This is because we believe that equipment charges are less likely to vary significantly between systems, whereas installation charges are more dependent on local labor and other costs that can be quite different in different communities. Third, operators may establish average charges only for similar types of equipment. Thus, for example, average charges may be established only for similar types of remotes or converters. Finally, when justifying equipment charges averaged across the operator's small systems, the operator must present a general description of the averaging methodology employed and a justification that it produces reasonable equipment rates. Based on the showing, local franchising authorities and the Commission may, for good cause, require that the operator set equipment rates in accordance with existing rules. We also note that in the Cost Proceeding, we are soliciting comment on what level of cost averaging we should require or permit operators to employ in setting rates. We may alter in that proceeding the cost averaging approach established today. We will additionally monitor the impact of our action today to assure that it does not harm subscribers. 220. As indicated, this administrative relief provides a methodology for setting unbundled equipment charges based on actual cost. As such, it will be available to all cable operators owning small systems. We see no reason to limit the eligibility for this small system relief to operators of a certain size. Moreover, this relief is not intended as an interim measure. Rather, operators may set equipment rates based on company-wide average costs subject, as indicated, to any decision on cost averaging the Commission may establish in the Cost Proceeding. 221. We believe that the 1992 Cable Act affords us sufficient discretion to adopt the dual approach described above, which is designed to "reduce administrative burdens and costs of compliance" for all systems that have 1,000 or fewer subscribers. The statute does not mandate that we provide the same level or magnitude of relief to every small system regardless of the resources or affiliation with a large operator. Notably, we have tailored the relief on the basis of our evaluation of the resources and capabilities of different types of small systems. We have balanced the public interest benefits of unbundling equipment and installation charges from programming charges against the administrative burdens on such systems. We find that varying degrees of relief are appropriate for small systems owned by different size operators, although every small system is afforded relief of some kind. Moreover, the fact that some members of Congress unsuccessfully objected to the provision regarding administrative relief for systems with less than 1,000 subscribers because MSO-owned systems would fall within this category, does not demonstrate that Congress intended to limit our flexibility in fashioning appropriate administrative relief. We thus conclude that our dual approach is consistent with the 1992 Cable Act. (iii) Other Proposals for Administrative Relief 222. Several petitioners, including Arizona Cable Television Association, et al. ("Arizona"), Mountain Cable ("Mountain"), Cable Services, and CATA, advocate complete exemption from rate regulation for all systems that are classified as small systems. As stated in the Rate Order, "[w]e do not believe that our responsibility under the Cable Act to ensure that consumers are protected from unreasonable rates permits us totally to exempt small systems ... from rate regulation." Id. at para 463. The language of the statute itself indicates that the Commission cannot entirely exempt small systems from rate regulation. First, as noted in the Commission's discussion in the Rate Order, Section 623(b)(1) of the Communications Act requires that the Commission's rules protect subscribers of "any" cable system not subject to effective competition. Thus, absent a statutory exemption, the regulatory requirements of the law apply to all systems, including small systems. Second, the statute expressly addresses special regulatory treatment of small systems in Section 623(i), but that Section authorizes the Commission only to "reduce" the administrative burdens and cost of compliance for small systems. Use of the term "reduce" implies that some amount of regulatory burden, although lessened, would remain. Similarly, because it is the "cost of compliance" that is to be reduced, it is apparent that some compliance is intended. If Congress had intended to exempt small systems entirely, or afford discretion to eliminate all burdens on such systems, it would have so stated. Accordingly, our efforts to reduce regulatory burdens and cost for small systems may not include complete exemption. 223. We also reject the "reasonable net revenue" test proposed by the Coalition and CATA. This test would exempt those systems with a net income margin of less than 15.5% from rate regulation. We believe that there is no dependable connection between this proposed test of profitability and a presumption that a small system's rates are reasonable. There are simply too many factors that can affect the gross revenue and net revenue calculations that are not necessarily related to reasonableness of rates. The expenses of the system may be unusually high for reasons that would normally not be acceptable in cost based rate regulation. For instance, the system may be using accelerated depreciation or the system may have unreasonable salary or other expenses paid to owners. Moreover, we foresee numerous practical problems in implementing such an approach. For example, we are concerned that efforts to verify the legitimacy of "net revenue" figures, to obtain additional information and to provide safeguards sufficient to maintain the integrity of a net revenue test would add such complexity as to defeat the purpose of regulatory simplification for small systems. Accordingly, we do not believe that this option is an appropriate way to reduce the regulatory burdens on small systems. 224. Arizona would require (rather than merely encourage) multiple franchise authorities regulating a single small system to jointly file for rate regulation certification. While we continue to encourage such joint certification, we do not believe that it is appropriate to override individual franchising authorities' prerogatives to make their own rate-setting determinations. 225. Finally, CATA proposes that we delay small system rate regulation pending a study of the effects of the benchmark, and that small systems be permitted to charge a rate within some percentage of the average national charge. We do not believe that the proposal to allow small systems to charge a rate within some percentage of an average national charge has merit. Such a proposal would merely create an alternative benchmark rate and would not necessarily be any simpler to administer. We also do not intend to delay rate regulation for small systems while we study the effect of our rate regulations on other systems. We note that we have already provided a transition mechanism for small operators. Moreover, by adopting the new rate- setting methodology and the average equipment cost approaches contained herein, we provide an easy, less burdensome means for small systems to comply with rate regulation. Any further delay in rate regulation for small systems -- other than those provided in our transition rules -- would be unnecessary. 2. Headend vs. Franchise Area Definition of Small Systems 226. In response to the Stay Order/Further Notice, many commenters argue that the definition of a small system should be changed from a "headend" to a "franchise area" basis. They contend that the franchise area definition would be consistent with the definition of cable system for other rate regulation purposes because regulated rates are determined on a franchise area basis. They also contend that the use of a franchise area definition will not place artificial barriers on the technical integration of different franchise areas with a single headend. Some commenters state that they have stopped such technical integration and consolidation because of the concern that they would lose their small system classification. 227. We continue to believe that determining small system size based on a system's principal headend, including any other headends or microwave receive sites that are technically integrated to the system's principal headend, best harmonizes our small system rule with most of our existing regulations on cable system size. For example, the existing 1,000 subscriber exemptions in the network non- duplication and public inspection file rules are based on a system's headend rather than franchise area. To use a franchise area definition would result in some segments of a single integrated cable operation receiving rate treatment different from other segments of the same operation. Moreover, if we changed the definition of small system to a franchise area basis, we would significantly increase the number of instances in which an operator would qualify for regulatory relief. However, many of these "small systems" would merely be franchise areas with 1,000 or fewer subscribers that were part of integrated cable systems serving much larger numbers of subscribers. Additionally, we continue to believe that the benefits from consolidated operation in a technically integrated headend sufficiently outweigh the benefits of classification as a small system so that future consolidation and improvements will not be unduly inhibited. 3. Termination of Rate Regulation Stay for Small Systems 228. As indicated, in the Stay Order/Further Notice, we stayed rate regulation of small systems until we further addressed small system issues on reconsideration. We have now completed our evaluation of small system issues. In addition, small systems have been afforded a substantial period of time to prepare for rate regulation, and we see no reason to delay for their subscribers the benefits of rate regulation. Therefore, we will terminate the stay for small systems as of May 15, 1994, the effective date of the rules adopted herein. Local authorities may provide initial notices of regulation to small systems as of that date, and the Commission will accept newly filed complaints concerning cable programming services tiers provided by small systems as of that date. Small systems must then submit a rate justification (or otherwise file a permitted response, such as a written notification that it intends to use the streamlined rate reduction process) within the 30 days prescribed in our rules. In addition, the statutory 180- day window for filing complaints concerning rates for cable programming services tiers in effect on May 15, 1994 will commence running on that date for small systems. III. Fourth Report and Order A. Introduction 229. In this Fourth Report and Order, we adopt a methodology for adjusting capped rates when channels are added to, or deleted from, a tier of regulated cable service. We also decline to modify our benchmark requirements to account for system upgrades initiated or completed shortly before the onset of rate regulation of cable service. 230. In our April, 1993 Rate Order, we provided that cable operators may pass through, after setting their initial regulated rates for a tier, any increases in programming costs for regulated services that exceeded inflation. The Rate Order did not, however, specify how the benchmark approach should be used to determine rates when channels were added or deleted to regulated service tiers. In the Third Further NPRM, we thus sought comment on how we should adjust capped rates to reflect channel additions or deletions. We now adopt in this decision a methodology for adjusting rates in these circumstances. This approach permits operators to adjust their rates by specified per-channel amounts that apply to channels that are added, dropped, or moved from one tier to another. As such, the methodology is an easy way to adjust capped rates that will facilitate programming changes. It also permits operators to recover fully programming expenses. B. Adjustments to Capped Rates for Addition and Deletion of Channels 1. Background 231. In the Third Further NPRM, we sought comment on what methodology should be adopted for applying the benchmark system to adjust capped rates when channels are added or deleted from regulated tiers. We sought comment first on the regulatory goals that should guide our adoption of a methodology for addition and deletion of channels. We tentatively concluded that the methodology we develop should achieve the objectives of protecting consumers from unreasonable rates while assuring the continued growth of the cable industry and the additional services that it can provide to subscribers. 232. We then solicited comment on three possible methodologies for adjusting capped rates when adding or deleting channels from a particular regulated tier. Under the first proposed method, the new charge for the tier would consist of the sum of: (1) the current permitted charge for the tier, and (2) a charge calculated by multiplying the benchmark rate by the number of new channels on the tier. Under this approach, the declining rate per channel reflected in the benchmark would be applied only to additional channels. The per channel charge for existing channels would not be adjusted downward to reflect the benchmark curve. We tentatively concluded that this approach should not be adopted because it would permit significantly higher rates; because it could not be readily used for setting rates when channels are deleted; and because it would permit tier pricing above the economies of scale observed in our Competitive Survey and reflected in the benchmark system. 233. Under the second methodology proposed in the Third Further NPRM, the new permitted rate for a regulated tier when channels are added or deleted would be the benchmark per-channel rate multiplied by the new number of channels on the tier. This approach would assure that the benchmark curve is fully reflected in rates because the charge for the tier would be calculated by applying the benchmark rate to all channels on the tier. 234. In addition, by applying the benchmark per channel rate to calculate new rates instead of the operator's permitted per channel rate as calculated using FCC Form 393, this approach would bring systems whose rates were above the benchmark down to the benchmark when they added or deleted channels. At the same time, however, it would allow systems whose regulated rates were below the benchmark to bring those rates up to the benchmark simply by adding or deleting channels from regulated service. We tentatively concluded that this approach should not be adopted because it would create substantial disincentives for cable operators with rates above the benchmark to add channels and because it could create undue incentives for systems with below benchmark rates to add channels, permitting substantially increased rates for such operators. 235. We tentatively concluded in the Third Further NPRM that we should adopt a third methodology, which we termed the "parallel track" approach. Under this approach, programming costs would be removed from an operator's permitted charge per tier. The remaining charge would then be adjusted to reflect the proportionate increase or decrease observed in the benchmark curve based on the new number of channels offered across all regulated tiers. The new level of programming expense for the tier would then be added back to the adjusted tier charge to obtain the new charge for the tier. We stated in the Third Further NPRM that this approach incorporates the downward sloping benchmark curve that we observe when operators in an unregulated environment added channels and thus passes on to subscribers the efficiencies and economies of scale reflected in the benchmark rates. 2. Regulatory Goals 236. Commenters expressing a view on this issue broadly agree with our proposed goals for the methodology for adjusting capped rates when channels are added or deleted from a regulated tier. They state, for example, that the Commission should seek to preserve incentives for cable operators to provide additional programming services to consumers, while protecting consumers from unreasonable rates. Several commenters suggested additional goals, such as furthering increased diversity and local origination of programming and promoting administrative ease. 237. We conclude that our methodology for adjusting capped rates when channels are added or deleted from regulated tiers should be consistent with, and further implement, our general approach for regulation of cable service rates. Thus, our regulations governing adjustments to capped rates should preserve the competitive rates produced by our requirements for setting initial regulated rates, since rates closer to competitive levels will best serve consumers. Moreover, lower rates for goods and services can in many cases increase the quantities demanded, and can further increase output as suppliers seek to meet that demand. Thus, maintaining rates at reasonable levels as channels are added or deleted from regulated cable service will benefit the public by making available to consumers a greater quantity and range of services at lower prices. 238. In addition, we believe that the cable industry can, and should, continue to grow and provide new and additional services to subscribers. In particular, operators should be given incentives to participate fully in the development of an advanced telecommunications infrastructure. Accordingly, a goal of our "going-forward" methodology is to allow cable operators to grow and develop new facilities and services, including new and innovative regulated programming services. This goal is fully consistent with the views of commenters that our methodology should promote a diversity of programming. Our methodology should also seek to further the statutory goal of reducing administrative burdens on subscribers, operators, and regulators. We explain below how the going-forward methodology we adopt today meets these stated goals. 3. The Going-Forward Methodology a. Comments 239. Several cable operators support the first alternative, arguing that it would provide the greatest incentives for operators to add channels because it permits the highest rates. All operators expressing a view oppose the second approach arguing that it would provide significant disincentives for systems above the benchmark to add additional programming to regulated tiers. Many operators support adopting the third alternative, the third alternative, provided certain modifications were made. Several operators, however, oppose it. They argue that the benchmark formula, and the underlying data on which it is based, are flawed and that, consequently, the formula cannot be incorporated into a methodology to set rates when channels are added or deleted from regulated tiers. 240. A number of programmers oppose the first alternative on the ground that it would discourage cable systems from purchasing expensive programming because, as described in the Third Further NPRM, it would not take into account actual programming costs. Programmers also strongly opposed the second option, arguing that it would provide significant disincentives for systems above the benchmark to add additional programming to regulated tiers. The majority of programmers expressing support for one of the alternatives favor the third alternative, although many proposed modifications while expressing their support. Programmers favor the third proposed methodology because in their view it reduces the disincentives for operators to purchase high quality programming contained in other approaches. 241. A number of franchising authorities also support the third alternative. Other franchising authorities oppose it because they believe that it would be difficult to administer. They support the second alternative because they view it as producing lower rates and because they believe it would be simpler to administer. A number of commenters favor permitting a mark-up on programming expense, arguing that this would encourage investment in new programming. b. Discussion 242. In addition to revealing a significant competitive differential that we will use to implement our revised benchmark approach, our Competitive Survey of industry rates as of September 30, 1992, established that, on average, charges per channel decrease as the number of channels offered by a system increases. This downward "curve" in per-channel rates may well reflect economies of scope and scale in the provision of regulated cable service. Using this curve in developing a going-forward methodology would benefit consumers by assuring that they receive the reduction in per channel rates that apparently arises as cable systems grow and add channels. At the same time, because the curve is based on observed per channel rates, we believe that a methodology that adjusts capped rates in accordance with this curve will permit operators to continue to invest in the provision of cable service. Thus, incorporation of this curve in our methodology will help achieve our goals of protecting consumers and permitting operators to respond to marketplace forces for provision of new services. 243. We emphasize that cable operators have failed to provide concrete cost or other factual information showing that the observed pattern of rates voluntarily established by the industry in an unregulated environment will not permit recovery of costs when new channels are added, including through facilities upgrades. For the most part, they merely offer generalized allegations that do no more than point out that the benchmark rate decreases on a per channel basis as the number of channels increases. Accordingly, we will adopt a methodology for adjusting capped rates that incorporates the downward curve of per- channel rates observed in our Competitive Survey. We also emphasize that those operators who believe that the rates determined under our new benchmark approach and the going- forward methodology we adopt, are inadequate when channels are added may make a cost-of-service showing in order to attempt to justify a higher rate. 244. The third alternative proposed in the Third Further NPRM is fully compatible with our revised benchmark formula and approach for setting regulated rates. For the same reasons that our statistical and other improvements to the benchmark formula better estimate the competitive differential, the revised benchmark formula more accurately captures the curve of declining per channel rates as the total number of offered regulated channels increases. Therefore, we can employ our new benchmark formula in a going-forward methodology, and it should permit operators to recover costs of providing additional channels while passing on to consumers the per-channel savings observed in our Competitive Survey. 245. Under the third alternative proposed in the Third Further NPRM, operators would subtract programming expenses from the permitted charge for each tier to establish a residual rate component for the tier. This is consistent with the requirements that we are adopting on further reconsideration for calculating all external costs and inflation adjustments. This treatment will achieve identical results as the method specified in the Rate Order but will be simpler to administer. At the same time, the operator will be able to fully recover in going-forward rate calculations the actual level of programming expense incurred. This approach will assure that operators may respond to demand for programming and recover their costs when adding channels. Accordingly, we believe that of the alternatives discussed in the Third Further NPRM, the third alternative will best achieve our goals for a methodology for applying the benchmark system to adjust capped rates when adding or deleting channels. We therefore adopt that approach. 246. In order to help assure that our methodology for adjusting capped rates when channels are added to, or deleted from, regulated tiers will help promote the growth and diversity of cable programming services, we will also permit operators a mark-up on new programming expense of 7.5%. This mark-up will apply only to any additional programming cost for a tier, measured on a per subscriber basis occurring after May 15, 1994. Operators must also reduce rates by any decreases in programming expense plus an additional 7.5% after that date. This will reduce incentives for operators to delete programming in order to replace it with new programming to which the mark-up could then be applied. 247. Under these requirements, operators will first remove all external costs from the tier charge and then adjust the residual component of the tier charge by a specified amount per channel when the total number of regulated channels increases. Should the total number of regulated channels decrease, the residual component of tier charge will be reduced by a specified amount. The per- channel adjustment factors used to calculate changes in permitted tier charges are derived from our benchmark equation and appear as a Table in the Technical Appendix. The Technical Appendix describes in some detail how they are computed. 248. When a cable system changes the number of regulated channels offered, it must average the initial and final number of channels and find the adjustment factor in the table corresponding to that average. For any service tier, the total permitted adjustment is the product of the per channel adjustment factor and the change in the number of regulated channels on that tier. The adjustment is positive if the number of regulated channels has increased and negative if the total number of regulated channels has decreased. If a cable operator is merely restructuring tiers and there is no change in the total number of regulated channels, then the operator would find its total number of regulated channels in the table, note the corresponding per channel adjustment factor, and calculate adjustments in network costs per tier as explained earlier in this paragraph. After the residual component of the tier charge is adjusted in this fashion, all external costs, including programming expenses, will be combined with the adjusted residual to determine the final tier charge. As stated, any increased level of programming expense will be entitled to a 7.5 percent mark-up. 249. The foregoing methodology for adjusting capped rates when channels are added or deleted from a regulated tier is set forth in detail in our new rule section 76.922(e). FCC Form 1210 and associated instructions also sets forth in detail this methodology for adjusting capped rates when channels are added to, or deleted from, a regulated tier, as well as for external cost and inflation adjustments generally. This methodology will provide a relatively simple way for operators to determine rates when new programming services are added to regulated offerings. It will thus facilitate the provision of new programming services, and is not unduly burdensome on operators and regulators. It is also fully consistent with our revised benchmark approach to setting initial regulated rates and can be used for deletions of channels, and moving channels between regulated tiers. This approach also assures that channel additions or deletions on one tier do not affect rates on other tiers. B. Upgrades Initiated Shortly Before Rate Regulation 1. Background 250. In the Third Further NPRM we sought comment on whether operators with rates below benchmark levels which initiated or completed system upgrades shortly before rate regulation should be permitted to raise rates to benchmark levels without any cost showing. We also solicited comments on alternatives to full cost-of-service showings that could permit recovery of such upgrade costs. In particular, we sought comment on whether the streamlined cost-of-service showing proposed in the Cost-of-Service NPRM should be applied to these situations. 2. Comments 251. Cable operators generally favored permitting operators that completed upgrades prior to regulation to raise their rates to the benchmark. A number of commenters supported permitting operators to use streamlined cost-of-service showings. One state commission, the Massachusetts Community Antenna Television Commission, questioned the extent to which rate increases are necessary to assure a reasonable return, since when systems are upgraded they may introduce at least some declining costs resulting from economies of scale as well as reduced maintenance costs. The state commission expressed concern regarding the possibility of a windfall, and pointed to cost-of-service proceedings as an option for systems that believe a reasonable return is denied by our other rate regulations. A coalition of local franchising authorities opposed permitting systems which initiated upgrades shortly before rate regulation from raising rates to the benchmark level because doing so, they contend, gives preferential treatment to operators depending on when they made improvements. They argue that there is no evidence before the Commission that an adjustment is necessary, given the way the benchmarks were calculated. GTE argued that these upgrades should not be granted external treatment because of the Commission's presumption that the initial rates cover system costs and that if rates are not adequate, the operator has the cost-of-service option. 3. Discussion 252. Because we have decided on reconsideration to replace the benchmark system with a requirement that, with certain exceptions, all rates be reduced by the competitive differential to avoid refund liability, it is no longer necessary or appropriate to address the issue raised in the Third Further NPRM as to whether operators with rates below benchmark levels which initiated or completed system upgrades shortly before rate regulation should be permitted to raise rates to benchmark levels without any cost showing. Moreover, even if we had retained the original benchmark approach for determining initial regulated rates we would decide not to permit operators with rates below benchmark levels which initiated or completed system upgrades shortly before rate regulation to raise rates to the benchmark level without any cost showing. As discussed in the Rate Order, below-benchmark rates are presumptively not unreasonably low for cable operators because they were voluntarily established by operators in an unregulated environment. Operators have not provided factual information that would alter the conclusion that rates voluntarily established by them as of the initial date of regulation will not be unreasonably low for them even if they have incurred upgrade costs. In addition, absent a cost showing, local franchising authorities and the Commission would not be assured that the benchmark rate reflects a reasonable cost- based rate for recovery of the costs of the upgrade. Also, it has not been shown how we could adequately define past upgrades to determine eligibility for this treatment. Accordingly, we conclude that we shall not permit operators to raise rates above otherwise permitted levels on account of upgrades initiated or completed before regulation without any cost showings. 253. Providing for a streamlined cost-of-service showing for past upgrades would require an identification of past upgrade costs and evaluation of them in accordance with appropriate cost-of-service standards. We believe that this would be an unusually complex undertaking. In addition, as discussed herein, it has not been shown that below-benchmark rates established by operators are unreasonably low for them. Accordingly, we will additionally not establish special streamlined cost-of-service showings for past upgrades. Operators for whom our general rate regulations do not permit adequate recovery of upgrades initiated or completed shortly before rate regulation took effect may file cost-of-service showings. IV. Fifth Notice of Proposed Rulemaking A. Termination of Transition Relief 254. As discussed above, we have determined that systems owned by small operators and systems with low prices will not have to apply the full 17 percent competitive differential pending our analysis of the relationship between costs and prices for those systems. We are initiating these cost studies in our Cost Proceeding. Accordingly, we are here providing notice that we will establish further requirements concerning permitted rates for systems currently eligible for transition treatment. As stated, depending on the results of our cost studies, these further provisions could require such systems to terminate transition relief and establish full reduction rates. B. Going-Forward Methodology 255. Cable operators are actively exploring new technical developments that may enable them to provide up to 500 channels. Some of these technical capabilities may involve significant modifications or additions to distribution plant. Others may involve compression and multiplexing techniques that permit derivation of many channels without significant new distribution plant. The benchmark table adopted in the Rate Order, and our table reflecting the efficiency curve observed in our Competitive Survey, establish per channel adjustments for systems with total channels on regulated tiers of 100 channels or less. It does not currently establish per channel rates for systems that provide more than 100 channels. 256. We solicit comment on whether we should establish a methodology for adjusting capped rates in situations where there are more than 100 regulated channels. We solicit comment generally on what that methodology should be. We also seek comment on whether we could use mathematical formulations derived from existing data or tables. We also solicit comment on whether, instead of adopting a methodology for setting rates for offerings of more than 100 channels, we should cap rates at the 100 channel level unless the operator could justify a higher rate based on a cost-of-service showing. We solicit comments on how any of these proposals would effect incentives for operators to provide additional channels on an "a la carte" basis. We additionally solicit comment on whether our going-forward methodology should be modified to provide greater or lesser compensation to operators for adjustments to capped rates when channels are added or deleted from regulated tiers, and whether this would better meet our goals of encouraging infrastructure development and growth of programming. Operators should provide a complete factual justification for any claims that the current methodology is inadequate. C. Commercial Rates 257. We have determined that we would not establish rules permitting special rates for regulated commercial cable service on reconsideration of the Rate Order. We stated, however, that allowances for commercial rates might help assure that rates for subscribers are reasonable if higher commercial earnings were offset by savings to consumers. Therefore, we solicit comment on whether we should establish regulations governing rates for regulated cable service provided to commercial establishments. In particular, we ask whether higher earnings for commercial establishments should be offset by lower rates to other subscribers. We solicit comment on whether the offset in rates to other subscribers should be exactly equal to the additional earnings from higher commercial rates. Alternatively, we could establish regulations that would mandate a specified level of sharing of earnings from higher commercial rates between operators and subscribers. We solicit comment on which approach would best serve subscribers and operators. We also solicit comment on what standards of reasonableness we could establish to govern commercial rates. V. Regulatory Flexibility Act Analysis A. Final Analysis for the Fourth Report and Order and Second Order on Reconsideration. 258. Pursuant to the Regulatory Flexibility Act of 1980, 5 U.S.C. ÀÀÀÀ 601-612, the Commission's final analysis with respect to the Fourth Report and Order and Second Order on Reconsideration is as follows: 259. Need and purpose of this action. The Commission, in compliance with ÀÀ 3 of the Cable Television Consumer Protection and Competition Act of 1992, 47 U.S.C. ÀÀ 543 (1992) pertaining to rate regulation, adopts revised rules and procedures intended to ensure cable subscribers of reasonable rates for cable services with minimum regulatory and administrative burden on cable entities. 260. Summary of issues raised by the public in response to the Initial Regulatory Flexibility Analysis. There were no comments submitted in response to the Initial Regulatory Flexibility Analysis. The Chief Counsel for Advocacy of the United States Small Business Administration (SBA) filed comments in the original rulemaking order. The Commission addressed the concerns raised by the Office of Advocacy in the Rate Order. 261. Significant alternatives considered and rejected. Petitioners representing cable interests and franchising authorities submitted several alternatives aimed at minimizing administrative burdens. In the present Order on Reconsideration, the Commission has attempted to accommodate the concerns addressed by these suggestions. For example, the Commission has chosen a more sophisticated economic model from among a number of statistical options to recalculate the competitive differential, and has reconsidered the benchmark approach such that all regulated cable systems will be required to establish rates based on the revised competitive differential. However, the Commission has determined that certain systems will not have to reduce rates by the full competitive differential immediately. Rather, the Commission will conduct cost studies of cable operators to allow systems with relatively low rates and operators with 15,000 or fewer subscribers to present evidence that the new competitive differential should not apply in full to them. These decisions will better ensure that regulated cable service rates are reasonable while reducing administrative burdens. In addition, the Commission provides administrative relief in the rate-setting process, and adopts simplified procedures concerning the requirements for calculating equipment costs and revenues for cable systems of 1,000 or fewer subscribers. 262. The Third Further NPRM in this proceeding presented three alternative methodologies for the adjustment of capped rates when channels are added or deleted from regulated service tiers. Many commenters supported, with some suggesting modifications, the approach the Commission tentatively endorsed in the Third Further NPRM. The Commission considered alternative methodologies and found on the basis of the record that the "parallel track" approach adopted in this Order, as well as the variety of revisions to its rate rules adopted here, will best achieve the goals of ensuring reasonable rates for consumers, promoting the growth and diversity of cable programming services, and facilitating ease of administration. B. Initial Regulatory Flexibility Analysis for the Fifth Notice of Proposed Rulemaking. 263. Pursuant to Section 603 of the Regulatory Flexibility Act, the Commission has prepared the following initial regulatory flexibility analysis (IRFA) of the expected impact of these proposed policies and rules on small entities. Written public comments are requested on the IRFA. These comments must be filed in accordance with the same filing deadlines as comments on the rest of the Notice, but they must have a separate and distinct heading designating them as responses to the regulatory flexibility analysis. The Secretary shall cause a copy of the Notice, including the initial regulatory flexibility analysis, to be sent to the Chief Counsel for Advocacy of the Small Business Administration in accordance with Section 603(a) of the Regulatory Flexibility Act, Pub. L. No. 96-354, 94 Stat. 1164, 5 U.S.C. Section 601 et seq. (1981). 264. Reason for action. The Cable Television Consumer Protection and Competition Act of 1992 requires the Commission to prescribe rules and regulations for determining reasonable rates for basic tier cable service and to establish criteria for identifying unreasonable rates for cable programming services. The Commission has adopted rate regulations that require a comparison to the rates of cable systems subject to effective competition, as defined in the Cable Act of 1992 and represented in the revised benchmark formula. This Notice proposes to establish regulations governing the setting of rates for regulated cable systems with more than 100 channels, and to consider separate rate regulations for commercial entities and rules for termination of transition relief. 265. Objectives. To propose rules to implement Section 3 of the Cable Television Consumer Protection and Competition Act of 1992. We also desire to adopt rules that will be easily interpreted and readily applicable and, whenever possible, minimize the regulatory burden on affected parties. 266. Legal Basis. Action as proposed for this rulemaking is contained in Sections 4(i), 4(j), 303(r) and 623 of the Communications Act of 1934, as amended. 267. Description, potential impact and number of small entities affected. We anticipate a possible impact on small entities because the Notice addresses the termination of transition relief for small systems owned by small operators. The Cable Act of 1992 defines a small system as serving 1,000 or fewer subscribers. 268. Reporting, record keeping and other compliance requirements. None. 269. Federal rules which overlap, duplicate or conflict with this rule. None. 270. Any significant alternatives minimizing impact on small entities and consistent with stated objectives. None. VI. Paperwork Reduction Act 271. The requirements adopted herein have been analyzed with respect to the Paperwork Reduction Act of 1980 and found to impose a new or modified information collection requirement on the public. Implementation of any new or modified requirement will be subject to approval by the Office of Management and Budget as prescribed by the Act. VII. Procedural Provisions 272. Ex parte Rules - Non-Restricted Proceeding. This is a non-restricted notice and comment rulemaking proceeding. Ex parte presentations are permitted, except during the Sunshine Agenda period, provided that they are disclosed as provided in Commission rules. See generally 47 C.F.R. Sections 1.1202, 1.1203, and 1.1206(a). 273. Pursuant to applicable procedures set forth in Sections 1.415 and 1.419 of the Commission's Rules, 47 C.F.R. Sections 1.415 and 1.419, interested parties may file comments on or before 45 days after publication in the Federal Register and reply comments on or before 75 days after publication in the Federal Register. To file formally in this proceeding, you must file an original plus four copies of all comments, reply comments, and supporting comments. If you want each Commissioner to receive a personal copy of your comments and reply comments, you must file an original plus nine copies. You should send comments and reply comments to Office of the Secretary, Federal Communications Commission, 1919 M Street, N.W. Washington, D.C. 20554. Comments and reply comments will be available for public inspection during regular business hours in the FCC Reference Center, Room 239, Federal Communications Commission, 1919 M Street N.W., Washington D.C. 20554. VIII. Ordering Clauses 274. Accordingly, IT IS ORDERED that, pursuant to Sections 4(i), 4(j), 303 (r), 612, 622(c) and 623 of the Communications Act of 1934, as amended, 47 U.S.C. ÀÀÀÀ 154(i), 154(j), 303(r), 532, 542(c) and 543 the rules, requirements and policies discussed in this Second Order on Reconsideration and Fourth Report and Order, ARE ADOPTED and Part 76 of the Commission's rules, 47 C.F.R. Part 76, IS AMENDED as set forth in Appendix A. 275. IT IS FURTHER ORDERED that, pursuant to Sections 4(i), 4(j), 303(r), 612(c), 622(c) and 623 of the Communications Act of 1934, 47 U.S.C. ÀÀÀÀ 154 (i), 154 (j), 303(r), 532 (c), 542(c), and 543, NOTICE IS HEREBY GIVEN of proposed amendments to Part 76, in accordance with the proposals, discussions, and statement of issues in this Notice of Proposed Rulemaking, and that COMMENT IS SOUGHT regarding such proposals, discussion, and statement of issues. 276. IT IS FURTHER ORDERED that, the Secretary shall send a copy of this Report and Order, including the Initial Regulatory Flexibility Analysis, to the Chief Counsel for Advocacy of the Small Business Administration in accordance with paragraph 603(a) of the Regulatory Flexibility Act. Pub. L. No. 96-354, 94 Stat. 1164, 5 U.S.C. ÀÀÀÀ 601 et seq. (1981). 277. IT IS FURTHER ORDERED that, the requirements and regulations established in this decision shall become effective May 15, 1994 with the exception of the 30 day notice requirement for rate changes to be codified at 47 C.F.R. Section 76.964(b) which shall be effective upon publication in the Federal Register. FEDERAL COMMUNICATIONS COMMISSION William F. Caton Acting Secretary APPENDIX A -- LIST OF COMMENTING PARTIES/PETITIONERS APPENDIX A MM Docket No. 92-266 Petitions for Reconsideration of Report and Order and Further Notice of Proposed Rulemaking in MM Docket No. 92- 266, 8 FCC Rcd 5631 (1993) Affiliated Regional Communications, Ltd. Alaska Cablevision, Inc. Alsea River Cable TV Arizona Cable Television Association, et. al. Atlanta Interfaith Broadcasters, Inc. Bank of New York Baraff, Koerner, Olender & Hochberg, P.C. Bell Atlantic Black Entertainment Television, Inc. Blade Communications, Inc. Booth American Company, et al. C-SPAN Cable Services Cablevision Systems Corporation California Cable Television Association Center for Media Education, et al. Century Communications Corp. Coalition of Small System Operators Colony Communications, Inc., et al. Comcast Cable Communications, Inc. Community Antenna Television Association, Inc. Community Broadcasters Association Continental Cablevision, Inc. Corning Incorporated; Scientific Atlanta, Inc. Crown Media, Inc. Discovery Communications, Inc. The Disney Channel E! Entertainment Television, Inc. Encore Media Corporation Fairmont Cable Harron Communications Corp. Higgins Lake Cable, Inc. Inland Bay Cable TV Associates InterMedia Partners King County, Wash., et al. Liberty Media Corp. Longview Cable Television Michigan C-TEC Communities Mountain Cablevision, Inc. Multichannel Communication Sciences, Inc. Municipal Franchising Authorities National Association of Telecommunications Officers and Advisors, et al. National Cable Television Association, Inc. Newhouse Broadcasting Corporation Northland Communications Corp. Paradise Television Network, Inc. Searle, Stanley M. SuperStar Connection Sur Corporation Tele-Communications, Inc. Time Warner Entertainment Company, L.P. TKR Cable Company/TKR Cable of Kentucky Turner Broadcasting System, Inc. Valuevision International, Inc. Viacom International, Inc. Video Data Systems Video Jukebox Network, Inc. Wometco Cable Corp. Comments/Oppositions to Petitions for Reconsideration Ad Hoc Rural Consortium Advanced Communications, Inc. Affiliated Regional Communications, Ltd. Arizona Cable Television Association Bell Atlantic BellSouth Telecommunications Bend Cable Communications, Inc., et al. Cable TV of Jersey City, Inc. Cablevision Industries Corporation, et al. Cablevision Systems Corporation Center for Media Education, et al. Consumer Electronics Group of the Electronic Industries Association Consumer Federation of America C-TEC Cable Systems Continental Cablevision, Inc. General Instrument Corporation GTE Service Corporation Home Recording Right Coalition Home Shopping Network, Inc. King County, et al. Liberty Cable Company, Inc. Medium-Sized Operators Group Michigan Communities National Association of Telecommunications Officers and Advisors, et al. National Association of Towns and Townships National Cable Television Association, Inc. National Telephone Cooperative Association Prevue Networks, Inc. Time Warner Entertainment Company, Inc. United States Telephone Association USA Networks Valuevision International, Inc. Viacom International, Inc. Videomaker Magazine Replies to Oppositions to Petitions for Reconsideration Cablevision Industries Corp., et al. Cablevision Systems Corporation Center for Media Education, et al. City of Saint Paul Coalition of Small System Operators Continental Cablevision, Inc. Corning Incorporated; Scientific-Atlanta, Inc. Discovery Communications, Inc. Engle Broadcasting King County, Wash., et al. Liberty Media Corporation Medium-Sized Operators Group Michigan C-TEC Corporation National Association of Telecommunications Officers and Advisors, et al. National Cable Television Association, Inc. Paradise Television Network, Inc. Puerto Rico Cable Television Association State of Hawaii Sur Corporation Televista Communications, Inc. Time Warner Entertainment Company, L.P. United Video, Inc. Valuevision International, Inc. Viacom International, Inc. Comments to Further Notice of Proposed Rulemaking in MM Docket No. 92-266, 8 FCC Rcd 5631 (1993) Alsea River Cable TV Arizona Cable Television Association, et al. Bell Atlantic Black Rock Cable TV Bonduel Cable TV, et al. Bye Cable, Inc. Cable Services Cableview Cablevision Industries Corporation, et al. Cascade Cable Systems City of Alexandria, VA Coalition of Small System Operators Colony Communications, Inc., et al. Community Antenna Television Association, Inc. Consumer Federation of America Continental Cablevision, Inc. Counsel to the Municipal Franchising Authorities Country Cablevision, Inc. Discovery Communications, Inc. Green River Cable TV Massachusetts Community Antenna Television Commission National Cable Television Association, Inc. National Association of Telecommunications Officers and Advisors, et al. Pacific Coast Cable Co. Stephen Cable TV, Inc. Tele-Communications, Inc. Time Warner Entertainment Company, L.P. Unites States Telephone Association Viacom International, Inc. Reply Comments to Further Notice of Proposed Rulemaking in MM Docket No. 92-266, 8 FCC Rcd 5631 (1993) Bell Atlantic Cablevision Industries Corporation, et al. Coalition of Small System Operators Consumer Federation of America Continental Cablevision, Inc. Maryland People's Counsel Mets Fans United Municipal Franchising Authorities State of New Jersey, Department of the Public Advocate Time Warner Entertainment Company, L.P. United Homeowners Association Comments to [Second] Further Notice of Proposed Rulemaking in MM Docket No. 92-266, 8 FCC Rcd 5585 (1993). Adelphia Communications Bend Cable Communications, Inc., et al. Coalition of Small System Operators Community Antenna Television Association Dennis Ready Falcon Cable Group GTE Service Corporation Medium-Sized Operators Group Mullin, Rhyne, Emmons and Topel, P.C. National Cable Television Association, Inc. National Association of Telecommunications Officers and Advisors, et al. National Telephone Cooperative Association Siskiyou Cablevision, Inc. Small Cable Business Association Tele-Media Corporation Union Telephone, Inc. Valley TV Cooperative, Inc. Wyoming Association of Municipalities Reply Comments to [Second] Further Notice of Proposed Rulemaking in MM Docket No. 92-266, 8 FCC Rcd 5585 (1993). Bell Atlantic Bend Cable Communications, Inc., et al. Coalition of Small System Operators GTE Service Corp. Petition for Reconsideration of First Order on Reconsideration, Second Report and Order and Third [Further] Notice of Proposed Rulemaking in MM Docket No. 92-266, 58 FR 46718 (September 2, 1993). New York Telephone Company and New England Telephone and Telegraph Company ("NYNEX") Attorney General of the State of Connecticut (Statement in support of NYNEX) Oppositions to Petitions for Reconsideration Cablevision Industries Corporation, et al. Continental Cablevision, Inc. Time Warner Entertainment Company, L.P. Viacom International, Inc. Reply to Oppositions to Petitions for Reconsideration NYNEX Comments to Third [Further] Notice of Proposed Rulemaking in MM Docket No. 92-266, 58 FR 46718 (September 2, 1993). Adelphia Communications Affiliated Regional Communications, Ltd. Austin, Texas, et al. Cable Television Association of New York, Inc. Cablevision Industries Corporation, et al. Cablevision Systems Corporation Community Antenna Television Association Continental Cablevision, Inc. Discovery Communications, Inc. et al. Falcon Cable TV, et al. GTE Service Corp. Hearst Corporation KBLCom, et al. Liberty Media Corporation Massachusetts Cable TV Media General Cable Municipal Franchising Authorities National Cable Television Association, Inc. National Association of Telecommunications Officers and Advisors, National Broadcasting Company New Jersey Board of Regulatory Authorities New York State Commission on Cable Television Newhouse Broadcasting Corporation Summitt Communications Tele-Communications, Inc. Tele-Media Corporation The Disney Channel Time Warner Entertainment Company, L.P. TKR Cable Company/TKR Cable of Kentucky Viacom International, Inc. Reply Comments to Third [Further] Notice of Proposed Rulemaking in MM Docket No. 92-266, 58 FR 46718 (September 2, 1993). Adelphia Communications Austin, Texas, et al. BellSouth Telecommunications Cablevision Industries Corporation, et al. Continental Cablevision, Inc. E! Entertainment Television, Inc. Fidelity Cablevision, Inc. GTE Service Corp. KBLCom, Inc., et al. Liberty Media Corporation Media General Cable National Cable Television Association, Inc. National Association of Telecommunications Officers and Advisors, et al. Time Warner Entertainment Company, L.P. Viacom International, Inc. APPENDIX B -- RULES APPENDIX B Title 47, Part 76 of the Code of Federal Regulations is amended as follows: PART 76 -- CABLE TELEVISION SERVICE 1. The authority citation for Part 76 continues to read as follows: Authority: Secs. 2, 3, 4, 301, 303, 307, 308, 309, 48 Stat. as amended, 1064, 1065, 1066, 1081, 1082, 1083, 1084, 1085, 1101; 47 U.S.C. Secs. 152, 153, 154, 301, 303, 307, 308, 309, 532, 535, 542, 543, 552 as amended, 106 Stat. 1460. 2. Section 76.901 is amended to revise paragraph (c) to read as follows: Section 76.901 Definitions * * * * * (c) Small System. A small system is a cable television system that serves 1,000 or fewer subscribers. The service area of a small system shall be determined by the number of subscribers that are served by the system's principal headend, including any other headends or microwave receive sites that are technically integrated to the principal headend. 3. Section 76.922 is amended to revise paragraphs (b), (c) and (d), and to add new paragraphs (e) and (f) to read as follows: Section 76.922 Rates for the basic service tier and cable programming services tiers. * * * * * (b) Permitted charge on May 15, 1994 (1) The permitted charge for a tier of regulated program service shall be, at the election of the cable system, either: (i) a rate determined pursuant to a cost-of- service showing; (ii) the full reduction rate; (iii) the transition rate, if the system is eligible for transition relief; or (iv) a rate based on a streamlined rate reduction, if the system is eligible to implement such a rate reduction. Except where noted, the term "rate" in this subsection means a rate measured on an average regulated revenue per subscriber basis. (2) Full reduction rate. The "full reduction rate" on May 15, 1994 is the system's September 30, 1992 rate, measured on an average regulated revenue per subscriber basis, reduced by 17 percent, and then adjusted for the following: (i) the establishment of permitted equipment rates as required by Section 76.923; (ii) inflation measured by the GNP-PI between October 1, 1992 and September 30, 1993; (iii) changes in the number of program channels subject to regulation that are offered on the system's program tiers between September 30, 1992 and the earlier of the initial date of regulation for any tier or February 28, 1994; and (iv) changes in external costs that have occurred between the earlier of the initial date of regulation for any tier or February 28, 1994, and March 31, 1994. (3) March 31, 1994 benchmark rate. The "March 31, 1994 benchmark rate" is the rate so designated using the calculations in Form 1200. (4) Transition rates. Systems owned by small operators and systems with low prices shall be eligible to establish a transition rate for a tier, pending a further order of the Commission. (A) Systems owned by small operators (i) For purposes of determining eligibility to establish a transition rate, a system owned by a small operator is a system owned by an operator that has a total subscriber base of 15,000 or fewer subscribers as of March 31, 1994. Systems owned by cable operators with between 15,000 and 16,000 subscribers may, upon a showing of substantial hardship, obtain a waiver from the Commission of the foregoing 15,000 subscriber limit. (ii) A system owned by a small operator shall not be eligible to establish a transition rate if the operator is owned or controlled by, or is under common control or affiliated with, a cable operator serving more than 15,000 subscribers. For purposes of this rule, a small cable operator will be considered affiliated with an operator serving more than 15,000 subscribers if such an operator holds a 20 percent or greater equity interest in the small operator. (iii) The transition rate for systems owned by small operators on May 15, 1994 shall be the system's March 31, 1994 rate, adjusted: (1) to establish permitted rates for equipment as required by Section 76.923 if such equipment rates have not already been established; and (2) for changes in external costs incurred between the earlier of the initial date of regulation for any tier or February 28, 1994, and March 31, 1994, to the extent such external cost changes are not already reflected in the system's March 31, 1994 rate. (B) Low-price systems (i) A low-price system is a system (1) whose March 31, 1994 rate is below its March 31, 1994 benchmark rate, or (2) whose March 31, 1994 rate is above its March 31, 1994 benchmark rate, but whose March 31, 1994 full reduction rate is below its March 31, 1994 benchmark rate, as defined in Section 76.922(b)(2), above. (ii) The transition rate on May 15, 1994 for a system whose March 31, 1994 rate is below its March 31, 1994 benchmark rate is the system's March 31, 1994 rate. The March 31, 1994 rate is in both cases adjusted: (1) to establish permitted rates for equipment as required by Section 76.923 if such rates have not already been established; and (2) for changes in external costs incurred between the earlier of initial date of regulation of any tier or February 28, 1994, and March 31, 1994, to the extent changes in such costs are not already reflected in the system's March 31, 1994 rate. The transition rate on May 15, 1994 for a system whose March 31, 1994 adjusted rate is above its March 31, 1994 benchmark rate, but whose March 31, 1994 full reduction rate is below its March 31, 1994 benchmark rate, is the March 31, 1994 benchmark rate, adjusted to establish permitted rates for equipment as required by Section 76.923 if such rates have not already been established. (C) Notwithstanding the foregoing, the transition rate for a tier shall be adjusted to reflect any determination by a local franchising authority and/or the Commission that the rate in effect on March 31, 1994 was higher (or lower) than that permitted under applicable Commission regulations. A filing reflecting the adjusted rate shall be submitted to all relevant authorities within 30 days after issuance of the local franchising authority and/or Commission determination. A system whose March 31, 1994 rate is determined by a local franchising authority or the Commission to be too high under the Commission's rate regulations in effect before May 15, 1994 will be subject to any refund liability that may accrue under those rules. In addition, the system will be liable for refund liability under the rules in effect on and after May 15, 1994. Such refund liability will be measured by the difference in the system's March 31, 1994 rate and its permitted March 31, 1994 rate as calculated under the Commission's rate regulations in effect before May 15, 1994. The refund liability will accrue according to the time periods set forth in Sections 76.942, and 76.961 of the Commission's rules. (5) Streamlined rate reductions. (A) Small systems that are not owned by or affiliated with any other system ("independent systems"), and small systems owned by small multiple system operators ("small MSOs"), that have not already restructured their rates to comply with the Commission's rules may establish rates for regulated program services and equipment by making a streamlined rate reduction. "Small MSOs" are those multiple system operators that (i) serve 250,000 or fewer total subscribers, (ii) own only systems with less than 10,000 subscribers each, and (iii) have an average system size of 1,000 or fewer subscribers. Independent small systems and small systems owned by small MSOs shall not be eligible for streamlined rate reductions if they are owned or controlled by, or are under common control or affiliated with, a cable operator that exceeds these subscriber limits. For purposes of this rule, a small system will be considered "affiliated with" such an operator if the operator holds a 20 percent or greater equity interest in the small system. (B) The streamlined rate for a tier on May 15, 1994 shall be the system's March 31, 1994 rate for the tier, reduced by 14 percent. A small system that elects to establish its rate for a tier by implementing this streamlined rate reduction must also reduce, at the same time, each billed item of regulated cable service, including equipment, by 14 percent. Regulated rates established using the streamlined rate reduction process shall remain in effect until: (1) adoption of a further order by the Commission establishing a schedule of average equipment costs; (2) the system increases its rates using the calculations and time periods set forth in FCC Form 1211; or (3) the system elects to establish permitted rates under another available option set forth in paragraph (b)(1) of this Section. (C) Implementation and notification. An eligible small system that elects to use the streamlined rate reduction process must implement the required rate reductions and provide written notice of such reductions to subscribers, the local franchising authority and the Commission according to the following schedule: (i) Where the franchising authority has been certified by the Commission to regulate the small system's basic service tier rates as of May 15, 1994, the system must notify the franchising authority and its subscribers in writing that it is electing to set its regulated rates by the streamlined rate reduction process. Such notice must be given by June 15, 1994, and must also describe the new rates that will result from the streamlined rate reduction process. Those rates must then be implemented within 30 days after the written notification has been provided to subscribers and the local franchising authority. (ii) Where the franchising authority has not been certified to regulate basic service tier rates by May 15, 1994, the small system must provide the written notice to subscribers and the franchising authority, described in subsection (i) above, within 30 days from the date it receives the initial notice of regulation from the franchising authority. The system must then implement the streamlined rate reductions within 30 days after the written notification has been provided to subscribers and the local franchise authority. (iii) Where the Commission is regulating the small system's basic service tier rates as of May 15, 1994, the system must notify the Commission and its subscribers in writing that it is electing to set its regulated rates by the streamlined rate reduction process. Such notice must be given by June 15, 1994, and must also describe the new rates that will result from the streamlined rate reduction process. Those rates must then be implemented within 30 days after the written notification has been provided to subscribers and the Commission. (iv) Where the Commission begins regulating basic service rates after May 15, 1994, the small system must provide the written notice to subscribers and the Commission, described in paragraph (iii) above, within 30 days from the date it receives an initial notice of regulation. The system must then implement the streamlined rate reductions within 30 days after the written notification has been provided to subscribers and the Commission. (v) If a complaint about its cable programming service rates has been filed with the Commission on or before May 15, 1994, the small system must provide the written notice described in paragraph (i), above, to subscribers, the local franchising authority and the Commission by June 15, 1994. If a cable programming services complaint is filed against the system after May 15, 1994, the system must provide the required written notice to subscribers, the local franchising authority or the Commission within 30 days after the complaint is filed. The system must then implement the streamlined rate reductions within 30 days after the written notification has been provided. (vi) A small system is required to give written notice of, and to implement, the rates that are produced by the streamlined rate reduction process only once. If a system has already provided notice of, and implemented, the streamlined rate reductions when a given tier becomes subject to regulation, it must report to the relevant regulator (either the franchising authority or the Commission) in writing within 30 days of becoming subject to regulation that it has already provided the required notice and implemented the required rate reductions. (6) Establishment of initial regulated rates. (A) Cable systems, other than those eligible for streamlined rate reductions, shall file FCC Forms 1200, 1205, and 1215 for a tier that is regulated on May 15, 1994 by June 15, 1994, or thirty days after the initial date of regulation for the tier. A system that becomes subject to regulation for the first time on or after July 1, 1994 shall also file FCC Form 1210 at the time it files FCC Forms 1200, 1205 and 1215. (B) A cable system will not incur refund liability under the Commission's rules governing regulated cable rates on and after May 15, 1994 if: (1) between March 31, 1994 and July 14, 1994, the system does not change the rate for, or restructure in any fashion, any program service or equipment offering that is subject to regulation under the 1992 Cable Act; and (2) the system establishes a permitted rate defined in paragraph (b) of this rule by July 14, 1994. The deferral of refund liability permitted by this subsection will terminate if, after March 31, 1994, the system changes any rate for, or restructures, any program service or equipment offering subject to regulation, and in all events will expire on July 14, 1994. Moreover, the deferral of refund liability permitted by this subsection does not apply to refund liability that occurs because the system's March 31, 1994 rates for program services and equipment subject to regulation are higher than the levels permitted under the Commission's rules in effect before May 15, 1994. (7) For purposes of this section, the initial date of regulation for the basic service tier shall be the date on which notice is given pursuant to ÀÀ 76.910 of our rules, that the provision of the basic service tier is subject to regulation. For a cable programming services tier, the initial date of regulation shall be the first date on which a complaint on the appropriate form is filed with the Commission concerning rates charged for the cable programming services tier. (8) For purposes of this section, rates in effect on the initial date of regulation or on September 30, 1992 shall be the rates charged to subscribers for service received on that date. (c) Subsequent permitted charge. The permitted charge for a tier after May 15, 1994 shall be, at the election of the cable system, either (1) a rate determined pursuant to a cost-of-service showing, or (2) a rate determined by application of the Commission's price cap requirements set forth in paragraph (d) below to a permitted rate determined in accordance with paragraph (b) above. (d) Price cap requirements. The Commission's price cap requirements allow a system to adjust its permitted charges for inflation and changes in external costs. After May 15, 1994, adjustments for changes in external costs shall be calculated by subtracting external costs from the system's permitted charge and making changes to that "external cost component" as necessary. The remaining charge, referred to as the "residual component," will be adjusted annually for inflation. Cable systems shall use FCC Form 1210 (or FCC Form 1211, where applicable) to justify changes in permitted rates made pursuant to the price cap requirements. (1) Calendar year quarters. All systems must use a calendar year quarter when adjusting rates under the price cap requirements. The first quarter shall run from January 1 through March 31 of the relevant year; the second quarter shall run from April 1 through June 30; the third quarter shall run from July 1 through September 30; and the fourth quarter shall run from October 1 through December 31. (2) Inflation adjustments. The residual component of a system's permitted charge may be adjusted annually for inflation. The annual inflation adjustment shall be based on inflation occurring from June 30 of the previous year to June 30 of the year in which the inflation adjustment is made, except that the first annual inflation adjustment shall cover inflation from September 30, 1993 until June 30 of the year in which the inflation adjustment is made. The adjustment may be made after September 30, but no later than August 31 of the next calendar year. Adjustments shall be based on changes in the Gross National Product Price Index as published by the Bureau of Economic Analysis of the United States Department of Commerce. Cable systems that establish a transition rate pursuant to paragraph (b)(4) above shall not be permitted to adjust rates on account of inflation until the transition rate adjusted for external costs and changes in numbers of regulated channels is less than, or equal to, the system's full reduction rate adjusted for inflation, external costs and changes in numbers of regulated channels. (3) External costs. (i) Permitted charges for a tier may be adjusted up to quarterly to reflect changes in external costs experienced by the cable system. In all events, a system must adjust its rates annually to reflect any decreases in external costs that have not previously been accounted for in the system's rates. A system must also adjust its rates annually to reflect any changes in external costs, inflation and the number of channels on regulated tiers that occurred during the year if the system wishes to have such changes reflected in its regulated rates. A system that does not adjust its permitted rates annually to account for these changes will not be permitted to increase its rates subsequently to reflect the changes. (ii) A system must adjust its rates in the next calendar year quarter for any decrease in programming costs that results from the deletion of a channel or channels from a regulated tier. (iii) Any rate increase made to reflect an increase in external costs must also fully account for all other changes in external costs, inflation and the number of channels on regulated tiers that occurred during the same period. Rate adjustments made to reflect changes in external costs shall be based on any changes in those external costs that occurred from the end of the last quarter for which an adjustment was previously made through the end of the quarter that has most recently closed preceding the filing of the FCC Form 1210 (or FCC Form 1211, where applicable). A system may adjust its rates after the close of a quarter to reflect changes in external costs that occurred during that quarter as soon as it has sufficient information to calculate the rate change. (iv) External costs shall consist of costs in the following categories: (1) state and local taxes applicable to the provision of cable television service; (2) franchise fees; (3) costs of complying with franchise requirements, including costs of providing public, educational, and governmental access channels as required by the franchising authority; (4) retransmission consent fees and copyright fees incurred for the carriage of broadcast signals; and (5) other programming costs. (v) The permitted charge for a regulated tier shall be adjusted on account of programming costs, copyright fees and retransmission consent fees only for the program channels or broadcast signals offered on that tier. (vi) The permitted charge shall not be adjusted for costs of retransmission consent fees or changes in those fees incurred prior to October 6, 1994. (vii) The starting date for adjustments on account of external costs for a tier of regulated programming service shall be the earlier of the initial date of regulation for any basic or cable service tier or February 28, 1994. (viii) Changes in franchise fees shall not result in an adjustment to permitted charges, but rather shall be calculated separately as part of the maximum monthly charge per subscriber for a tier of regulated programming service. (ix) Adjustments to permitted charges to reflect changes in the costs of programming purchased from affiliated programmers, as defined in Section 76.901 of these rules, shall be permitted as long as the price charged to the affiliated system reflects either prevailing company prices offered in the marketplace to third parties (where the affiliated program supplier has established such prices) or the fair market value of the programming. (x) Adjustments to permitted charges on account of increases in costs of programming shall be further adjusted to reflect any revenues received by the operator from the programmer. (xi) In calculating programming expense, operators may add add a mark-up of 7.5% for new programming added after May 15, 1994 and shall reduce rates by decreases in programming expense plus an additional 7.5% for decreases occurring after May 15, 1994. (e) Changes in the number of channels on regulated tiers. A system may adjust the residual component of its permitted rate for a tier to reflect changes in the number of channels offered on the tier on a quarterly basis. Cable systems shall use FCC Form 1210 (or FCC Form 1211, where applicable) to justify rate changes made on account on changes in the number of channels on a regulated tier. Such rate adjustments shall be based on any changes in the number of regulated channels that occurred from the end of the last quarter for which an adjustment was previously made through the end of the quarter that has most recently closed preceding the filing of the FCC Form 1210 (or FCC Form 1211, where applicable). However, when it deletes channels in a calendar quarter, a system must adjust the residual component of the tier charge in the next calendar quarter to reflect that deletion. The following table shall be used to adjust permitted rates for a tier for changes in the number of channels offered on the tier. The entries in the table provide the cents per channel per subscriber per month by which cable operators will adjust the residual component using FCC Form 1210 (or FCC Form 1211, where applicable). Average Number of Regulated Channels Per-Channel Adjustment Factor Average Number of Regulated Channels Per-Channel Adjustment Factor 7 $0.52 14 0.14 7.5 0.45 14.5 0.13 8 0.40 15-15.5 0.12 8.5 0.36 16 0.11 9 0.33 16.5-17 0.10 9.5 0.29 17.5-18 0.09 10 0.27 18.5-19 0.08 10.5 0.24 19.5-21.5 0.07 11 0.22 22-23.5 0.06 11.5 0.20 24-26 0.05 12 0.19 26.5-29.5 0.04 12.5 0.17 30-35.5 0.03 13 0.16 36-46 0.02 13.5 0.15 46.5 and above 0.01 In order to adjust the residual component of the tier charge when there is a change in the number of channels on a tier, the operator shall perform the following calculations: (1) take the sum of the old total number of channels on tiers subject to regulation (i.e., tiers that are, or could be, regulated) and the new total number of channels and divide the resulting number by two; (2) consult the above table to find the applicable per channel adjustment factor for the number of channels produced by the calculations in step (1). For each tier for which there has been a change in the number of channels multiply the per-channel adjustment factor times the change in the number of channels on that tier. The result is the total adjustment for that tier. It is positive if the number of channels on the tier has increased and negative if the number of channels has decreased. (f) Permitted charges for a tier shall be determined in accordance with forms and associated instructions established by the Commission. 4. Section 76.923 is amended to add paragraph (l) to read as follows: Section 76.923 Rates for equipment and installation used to receive the basic service tier * * * * * (l) Company-wide averaging of equipment costs. For the purpose of developing unbundled equipment charges as required by Section 76.923(b), a cable operator may average the equipment costs of its small systems at any level, or several levels, within its operations. This company-wide averaging applies only to an operator's small systems as defined in Section 76.901(c) above; is permitted only for equipment charges, not installation charges; and may be established only for similar types of equipment. When submitting its equipment costs based on average charges to the local franchising authority or the Commission, an operator that elects company-wide averaging of equipment costs must provide a general description of the averaging methodology employed and a justification that its averaging methodology produces reasonable equipment rates. The local authority or the Commission may require the operator to set equipment rates based on the operator's level of averaging in effect on April 3, 1993, as required by Section 76.924(d). 5. Section 76.934 is amended to redesignate current Section 76.934 as paragraph (a) and to add new paragraphs (b) and (c). Section 76.934 Small Systems (a) A franchising authority that has been certified, pursuant to ÀÀ 76.910, to regulate rates for basic service and associated equipment may permit a small system as defined in ÀÀ 76.901 to certify that the small system's rates for basic service and associated equipment comply with ÀÀ 76.922, the Commission's substantive rate regulations. (b) Initial regulation of small systems (1) If certified by the Commission, a local franchising authority may provide an initial notice of regulation to a small system, as defined by Section 76.901(c), on May 15, 1994. Any initial notice of regulation issued by a certified local franchising authority prior to May 15, 1994 shall be considered as having been issued on May 15, 1994. (2) The Commission will accept complaints concerning the rates for cable programming service tiers provided by small systems on or after May 15, 1994. Any complaints filed with the Commission about the rates for a cable programming service tier provided by a small system prior to May 15, 1994 shall be considered as having been filed on May 15, 1994. (3) A small system that receives an initial notice of regulation from its local franchising authority, or a complaint filed with the Commission for its cable programming service tier, must respond within the time periods prescribed in Sections 76.930 and 76.956. (c) Statutory period for filing initial complaint. A complaint concerning a rate for cable programming service or associated equipment provided by a small system that was in effect on May 15, 1994 must be filed within 180 days from May 15, 1994. (d) Petitions for extension of time. Small systems may obtain an extension of time to establish compliance with rate regulations provided they can demonstrate that timely compliance would result in severe economic hardship. Requests for extension of time should be addressed to the local franchising authority concerning basic service and equipment rates and to the Commission concerning rates for a cable programming service tier and associated equipment. The filing of a request for an extension of time to comply with the rate regulations will not toll the effective date of rate regulation for small systems or alter refund liability for rates that exceed permitted levels after May 15, 1994. 6. Section 76.952(a) is amended to revise paragraph (a) to read as follows: Section 76.952 Information to be provided by cable operator on monthly subscriber bills. (a) The name, mailing address and phone number of the local franchising authority and the Cable Services Bureau of this Commission. * * * * * 7. Section 76.953(a) is amended to revise paragraph (a) to read as follows: (a) Complaint regarding a rate in effect on September 1, 1993. Notwithstanding paragraph (b) of this section, a complaint regarding a rate for cable programming service or associated equipment in effect on September 1, 1993, must be filed by February 28, 1994, except as provided in Section 76.934(c) with respect to small systems. * * * * * 8. Section 76.958 is added to Subpart N to read as follows: Section 76.958 Notice to Commission of rate change while complaint pending A regulated cable operator that proposes to change any rate while a cable service tier complaint is pending before the Commission shall provide the Commission at least 30 days notice of the proposed change. 9. Section 76.964 is amended to redesignate current Section 76.964 as paragraph (a) and to add new paragraph (b). Section 76.964 Notices to subscribers (a) In addition to the requirement of Section 76.309(c)(3)(i)(B) regarding advance notification to customers of any changes in rates, programming services or channel positions, a cable operator shall give the relevant franchising authority a minimum of 30 days advance written notification of any changes in rates for cable programming services or associated equipment. (b) Cable systems shall give 30 days written notice to both subscribers and local franchising authorities before implementing any rate or service change. Such notice shall state the precise amount of any rate change and briefly explain in readily understandable fashion the cause of the rate change (e.g., inflation, changes in external costs or the addition/deletion of channels). When the change involves the addition or deletion of channels, each channel added or deleted must be separately identified. Notices to subscribers shall inform them of their right to file complaints about changes in cable programming service tier rates and services with this Commission within 45 days of the rate or service change being reflected in their bill, and shall provide the address and phone number of both the local franchising authority and the Cable Services Bureau of this Commission. 10. A new Section 76.986 is added to Subpart N to read as follows: Section 76.986 "A la carte" offerings (a) Collective offerings of unregulated per-channel or per- program ("a la carte") video programming shall not be regulated if: (1) the price for the combined package does not exceed the sum of the individual charges for each component of service, and (2) the cable operator continues to provide the component parts of the package to subscribers separately in addition to the collective offering. The second condition will be met only when the per channel offering provides consumers with a realistic service choice. Collective offerings available on April 1, 1993 shall not be regulated if subsequently offered on the same terms and conditions as were in effect on that date. (b) In reviewing a basic service rate filing, local franchising authorities may make an initial decision addressing whether a collective offering of "a la carte" channels will be treated as an unregulated service or a regulated tier. The franchising authority must make this initial decision within the 30 day period established for review of basic cable rates and equipment costs in Section 76.933(a), or within the first 60 days of an extended 120 day period (if the franchise authority has requested an additional 90 days) pursuant to Section 76.933(b). The franchising authority shall provide notice of its decision to the cable system and shall provide public notice of its initial decision within seven days pursuant to local procedural rules for public notice. Operators or consumers may make an interlocutory appeal of the initial decision to the Commission within 14 days of the initial decision. Operators shall provide notice to franchise authorities of their decision whether or not to appeal to the Commission within this period. Consumers shall provide notice to franchise authorities of their decision to appeal to the Commission within this period. (c) A limited initial decision under paragraph (b) shall toll the time periods under Section 76.933 within which local authorities must decide local rate cases. The time period shall resume running seven days after the Commission decides the interlocutory appeal, or seven days following the expiration of the period in which an interlocutory appeal pursuant to paragraph (b) may be filed. (d) A local franchising authority alternatively may decide whether a collective offering of "a la carte" channels will be treated as an unregulated service or a cable programming services tier as part of its final decision setting rates for the basic service tier. That decision may then be appealed to the Commission as provided for under Section 76.945. APPENDIX C -- TECHNICAL APPENDIX APPENDIX C TECHNICAL APPENDIX Since the original benchmark was adopted in the Report and Order, we have made substantial improvements in the methodology for determining competitive cable rates and adjusting to reasonable levels the rates of cable systems subject to regulation. The Commission has improved both the quality of the survey data base and the approach to estimation of the competitive differential. These revisions are based on additional staff analysis, as well as suggestions from commenters. This Appendix reports on data collection and statistical activities that have occurred since the Report and Order. Data Adjustments Several commenters identified missing data and errors in the Cable Rate Survey data base that potentially affected the results of the analysis. The staff attempted to improve the quality of the data base by incorporating additional information received after the original data base was constructed. The National Cable Television Association (NCTA) submitted a data diskette with corrections to the data base, based on phone inquiries to cable systems. The staff randomly checked the changes and determined that the original data were faulty and that the new values fell within a plausible range. The means of variables that contained changes were compared with their original values to confirm that no major changes in the data resulted. Then the NCTA revisions were incorporated into the data base. Other commenters pointed out specific errors in the data, and some cable systems sent in amended forms. The staff also detected numerous errors in the data. Checks were made to see that values of important variables lay within possible ranges (e.g., that the number of homes passed was greater than or equal to the number of subscribers). For important variables, data points that appeared to be either implausibly high or implausibly low were pinpointed. The variables given particular scrutiny included monthly charge, total number of channels, total number of satellite channels, rate per channel, and number of tier subscribers. Cases that were outliers in the benchmark regressions were also examined. The classification of competitive systems was examined with particular care. Where evidence of data errors appeared from any of these sources, the staff called the cable system to obtain the correct information, which was then incorporated into the data base. Some of the data amendments made it necessary to change the sample used in the analysis. Survey forms that were received after the initial deadline were edited, coded, and entered into the data base. Eleven new cases were added. Systems that had been in the original data base but that carried only broadcast signals were removed from the data base because they did not meet the statutory definition of a cable system. Five such records were deleted. One record was deleted because the system had just begun operation and had only part-year data. One record was deleted because the system served only commercial accounts. Three groups of cable systems--low penetration systems, overbuilds, and municipals--are subject to "effective competition" as defined in Section 623(l)(1) of the Communications Act of 1934, 47 U.S.C. 543(l)(1). The staff determined membership in the low-penetration portion of the competitive sample by calculating penetration as a percentage of homes in the franchise area (the statutory definition) rather than depending on the systems' reporting of their penetration. Recalculating this variable resulted in the addition of 134 franchise areas to the low- penetration sample and the removal of 4 franchise areas. We believe that much of the increase in the number of franchise areas reported as having low penetration occurred because many systems reported penetration as a percentage of homes passed (the industry convention) rather than of franchise area homes (the statutory definition). We believe that the new sample definition is a more accurate embodiment of the statutory definition than is the old one. Franchise areas that belonged to the overbuild and municipal samples were also included in the low-penetration sample if they met the statutory criterion. Thus the coefficient of the low- penetration dummy variable is interpreted as the effect of low penetration on revenue per subscriber separately from any effect of overbuild competition or municipal status. In the overbuild sample, 4 franchises were found to have been coded into the wrong competitive classification. Multichannel competitors existed in these franchise areas but failed to meet 1 of the 2 statutory criteria for overbuild competition--50 percent of franchise area homes passed and 15 percent aggregate franchise area penetration by multichannel competitors other than the largest one. Two of these franchise areas were reclassified as low- penetration systems; the other 2 were reclassified as noncompetitive. One new overbuild system was added to the data base. In the municipal sample, the competitive status of 4 franchise areas was misreported. Two of them were reclassified as overbuilds and 2 were reclassified as noncompetitive. The systems reclassified as noncompetitive were retained in the sample if they belonged in the randomly-chosen noncompetitive sample; otherwise they were dropped from the analysis. The net result was 1 less overbuild system and 4 fewer municipal systems. As in the original benchmark analysis, not all of the observations collected were used in the analysis. Data for all systems whose responses to the questionnaires and telephone inquiries established that they met the statutory definition of effective competition were included. Data from a noncompetitive sample consisting of first franchise areas of systems in the random sample that did not face effective competition also were included. Data for systems that were sent questionnaires only because they were believed likely to face effective competition, but subsequently proved noncompetitive, were excluded; data for second franchise areas of systems in the random sample were excluded because they were not independent observations; and data from the sample of 100 largest cable systems were not used unless they also belonged to the random noncompetitive sample or faced effective competition. The sample used in current analyses consists of 496 records, including 253 noncompetitive franchise areas, 205 in the low-penetration sample, 45 in the overbuild sample and 12 in the municipal sample. Of the low-penetration franchise areas, 14 also belong to the overbuild sample and 4 belong to the municipal sample. Some observations dropped out of the analysis sample in the regression analysis because of missing data. The sample used in the final regression consists of 420 records, of which 237 were in the noncompetitive sample, 148 in the low-penetration sample, 39 in the overbuild sample, and 11 in the municipal sample. Of the low-penetration franchise areas in this sample, 11 belong to the overbuild sample and 4 belong to the municipal sample. This sample includes 74 more low-penetration franchise areas and 8 more overbuilds, but 4 fewer municipal systems, than the sample in the original benchmark regression. There were more overbuilds in the regression sample despite a net reduction in the total number of overbuilds because corrections in the data caused fewer observations to be unusable because of missing data. Disaggregation of the Competitive Sample In the Report and Order, we noted that the three competitive samples that we identified based on the statute- -overbuilds, municipals, and low penetration--had very different price behavior. In the benchmark regression, using a single dummy variable for all three samples produced a coefficient of -0.094, but when the low penetration systems were excluded from the sample, the differential increased in magnitude to -.28. When we used separate dummy variables for each of the three competitive samples, we found that the coefficient of the variable for the low penetration sample was positive, albeit small, and statistically insignificant, while the other 2 competitive dummy variables had negative and statistically significant coefficients. Because the differences among the coefficients of the dummies for the three separate competitive samples were statistically significant, we concluded that it was inappropriate to lump them together and represent them with a single variable in the regression analysis. Therefore, it was decided early on to estimate separate coefficients for each of the three groups. Hazlett argued that the low penetration observations should be dropped because they would distort the analysis. However, our disaggregated approach to estimating the competitive differential is robust to their inclusion or exclusion. We examined carefully each of the three samples defined as "effectively competitive." In some cases the variables were refined to improve their ability to represent actual economic competition. Low penetration sample. While low penetration may be associated with sharing a market with a competitor, it may also result from other conditions unrelated to competition. A system that has just begun operation or a system that serves a low-income neighborhood may have low franchise area penetration as a consequence. Low penetration may also result from high prices or poor service. Further, the statutory definition of low penetration is based on homes in the franchise area, not homes passed by the system. Yet cable systems frequently offer service only to a portion of the franchise area, often because the remainder of the franchise area is sparsely populated or for some other reason too costly to serve. Thus low penetration may reflect the geographic limits of the system rather than competition or anything related to system service or choices of potential subscribers. In fact, of the low-penetration systems in our analysis sample, 37 percent have more than 30 percent penetration of homes passed. For these reasons, low penetration, as defined in the statute, appears unlikely to be a useful indicator of competitive behavior. Low-penetration systems might have higher prices than equally competitive systems if they faced systematically higher costs. One cause of high costs per subscriber might be low demand, requiring systems to spread fixed costs over a smaller number of subscribers. As described below, we estimated a variety of equations including several cost and demand proxies. The low-penetration coefficient remained statistically insignificant. Overbuild sample. Next we attempted to refine the measurement of competition for the augmented overbuild sample. Many systems face competition in only a portion of their service areas. We hypothesized that the intensity of competition a system faces in a franchise area varies with the amount of actual overlap with a competitor, that is, the fraction of households the system passes that are actually passed by a competitor. While the data do not permit us to calculate this proportion precisely, we constructed a variable that forms a lower bound for it. This overlap variable takes on a value between zero and one for all observations in the augmented overbuild sample and takes on a value of zero for all other observations. The overlap variable described above measures overlap within a franchise area. Many systems, however, charge uniform prices in all franchise areas. In these cases, the appropriate geographic area for the measurement of competitive overlap is the system service area, not the franchise area. The distinction is empirically important in cases where a system with many franchise areas faces competition in only one or a few franchise areas. In these cases, if the degree of overlap in franchise areas with competition is taken as the measure of competition, the extent of competition to the system will be overstated because the diluting effects of franchise areas without competition will not be taken into account. This mismeasurement will cause the effect of competition on price to be understated, since the observed price response will have been caused by a smaller proportion of homes with competition than the measured proportion. This effect is likely to be greater in large than in small systems because if only one or a few of many franchise areas is competitive it is likely to constitute a smaller proportion of the whole system. Understating the effect of competition on price for large systems might introduce an erroneous appearance that the effect varies with system size. To account for this effect, for overbuild systems (including municipal ones) we estimated the percentage of homes passed in the entire system service area that are also passed by a competitor. Again, the data permit only an estimate, not an exact calculation. We constructed a combined overlap variable (designated OVL) that uses the system overlap measure for overbuild systems that report that they charge the same prices in all franchise areas, and uses the franchise area overlap measure for other overbuild systems. The variable takes on the value of zero for non- overbuild systems. If any two of the three overbuild variables (i.e., franchise overlap, OVL, and B) were included in the regression, none had a significant coefficient due to multicollinearity between the variables. The system overlap variable OVL resulted in a better fit than the franchise area overlap variable. Of the three taken individually, the franchise overlap variable resulted in the worst fit. When B was included and OVL was excluded, the regression had as good a fit as the one where OVL was included and B was excluded. Thus it appears that statistically either B or OVL is an equally good measure of the competitive effect. We chose OVL because we believe it captures the effect of head-to-head competition better than a simple zero-one dummy variable. Therefore, we selected OVL as the best measure of competition. OVL is generally positive for both the B (overbuild) and C (municipal) samples, but not the A1 (low penetration) sample. Therefore, when OVL is included along with B and/or C, the coefficients of B and C reflect the price deviation from the level estimated for the degree of competition measured by OVL. When OVL is excluded from the regression, the coefficients of B and C reflect the price deviation from the level estimated for no competition. Municipal sample. The effects of municipal ownership on the pricing behavior of cable systems is unclear a priori. City governments might choose to subsidize cable service, directly or indirectly, as a service to residents, causing municipal systems to have low costs and prices relative to private ones; municipal governments might view cable service as a source of revenues to subsidize other services; or, finally, municipal governments might view cable systems as neither a source of subsidies nor as a service to subsidize. The municipal sample in the survey consists of pairs of systems, a municipally-owned system and a privately-owned competitor, serving the same area. In order to test the possibility that municipally-owned systems have lower costs or different objectives than their private rivals, we constructed separate dummy variables for the 2 municipal subsamples. The difference between their coefficients was not statistically significant, so we retained a single municipal dummy variable. All but 2 of the systems in our municipal sample met the overbuild criterion as well. Consequently we included the overbuild municipal systems in the overbuild sample, creating an augmented overbuild dummy variable. This variable had a value of 1 for franchises in the augmented overbuild sample and 0 otherwise. In subsequent analyses, we used a variable representing overbuild status and a separate variable representing the additional effect of belonging to the municipal sample. Equations Estimated for Our New Model Using the random sample and the samples of overbuilds, municipals, and low penetration franchises, we estimated various equations using ordinary least squares multiple regression analysis. The use of multiple regression analysis permits the estimation of the separate impact of each variable included in the regression, holding the other variables constant. The main purpose of the regressions was to estimate the impact of competition on rates. Therefore, the variables that were of primary concern were those designed to measure the competitive effect. However, many other characteristics of systems and their markets affect rates. If the effects of these factors are not accounted for, and if they differ between competitive and non- competitive systems, they may bias the estimates of the variables measuring the effects of competition. Accordingly, an attempt was made to include in the model as many other variables as one could reasonably argue to be relevant and were found to have coefficients that were statistically significantly different from zero. Functional Form Both linear and logarithmic forms of the dependent variable were tried. The choices of the independent variables that resulted in the best fit were the same regardless of which form of the dependent variable was used. In terms of overall statistical fit, there was little to choose between a linear and logarithmic dependent variable; the difference between the fits was insignificant. In light of this statistical insignificance, we considered the underlying economic situation being analyzed. The logarithmic form assumes constant percentage effects of changes in the independent variables, which can be estimated using the coefficients. This implies that the competitive effect is a constant percentage reduction in rates. On the other hand, the linear form implies that the competitive effect is a constant dollars and cents reduction in rates. We were not aware of a plausible economic model that would produce price-margins with constant absolute markups across systems. To the extent that higher rates are indicative of higher price-cost margins, the constant percentage differential is more appropriate. Accordingly, we chose the logarithmic form for conducting most of the analysis. In the logarithmic form that was used, we took natural logarithms of only some of those variables whose values were always positive. The size variables, measuring the number of system subscribers and the number of channels, were found to give a better fit as reciprocals (which have values of between 0 and 1). In addition, logarithms were not taken of dummy variables (whose value is either 0 or 1) or proportions (whose value could be 0). Thus the general form that was used was where ln denotes the natural logarithm, y is the dependent variable, the x's are the positive independent variables, the p's are the dummy variables, proportions, and reciprocals, and u is the random error term.y = e SUP a x SUB 1 SUP b SUB 1 x SUB 2 SUP b SUB 2 ... e SUP { c SUB 1 p SUB 1 } e SUP { c SUB 2 p SUB 2 } ... e SUP u . The estimated coefficients are a, which is the intercept, the b's, which can be interpreted as elasticities for the x's (the percentage change in y associated with a one percent increase in the x value), and the c's, from which we can compute the proportionate impactof a change in from 0 to 1 using the formula , where e is the base of the natural logarithms.p SUB ie SUP { c SUB i (g - f) } - 1.c SUB i Variables Used The dependent variable used in the regression analysis was the average franchise area revenues from regulated services per subscriber per month, expressed in logarithmic form. This was used, instead of a quoted rate for basic service, because there may have been different amounts and directions of cross-subsidies among services for different cable companies. The dependent variable was recalculated to improve the measurement of equipment and installation revenue and franchise fees. The estimates in cases of missing data were carefully checked. The staff had information suggesting that cable systems rarely charge full list price for installations, which would cause the calculated value of installation revenue to overstate the true value. To correct for this bias, franchise area installation revenue was estimated from system installation revenue, multiplied by the proportion of system subscribers in the franchise area. The franchise fee calculation was also made slightly more precise by including equipment revenue in the base on which franchise fees are calculated when franchise fees are based on total subscriber revenue and not itemized. The revenue variable was then recalculated on a per- subscriber, rather than subscriber-channel, basis. The independent variables in the regression can be grouped into the following categories: competitive variables, size variables, product mix variables, and other cost variables. Various different variables were tested for inclusion in each of these categories. The ones that were included in our final equation are discussed here. Others that were tried and eliminated (generally because their estimated coefficients were not significantly different from zero) are discussed in the next section. The key variable used to measure the effect of competition is overlap, the proportion of the service area that is also served by a competitor. This variable can take any value from 0 to 1, with 0 representing no competition and 1 representing competition in the entire service area. The version of this variable that gave the best fit, labeled OVL, uses the system area as the service area for those franchises that reported that the same rates were charged for all franchises of their system, and uses the franchise area as the service area if the rates differ among the franchises of their system (or if there is only 1 franchise in the system). The coefficient of this variable is the primary basis for the measure used to determine the price differential that is deemed to be due to competition, other things being equal. A second variable reflecting membership in a group defined by the statute as competitive is a dummy variable for low penetration. This variable is identified as A1. There are three other variables that might be considered to be a hybrid of competitive and other cost variables. One of these is C, which is a dummy variable which is 1 for municipally owned franchises or privately owned franchises competing against municipally owned franchises. It could be argued that municipally owned franchises have lower costs due to direct or indirect government subsidies or other factors. On the other hand, as noted below, it appears that the privately owned franchises may be charging rates that are even lower than the model would otherwise predict compared to the municipally owned ones they compete against, which would tend to support the argument that the coefficient of this variable measures the impact of more intense price competition than is picked up by the overlap variable. As an indication of this, the length of time that the municipals have been subject to competition has on average been less than that of the overbuild sample, and it is generally accepted economic theory that price competition tends to be more intense in the early years of competition. The other 2 hybrid variables are related to whether the system is connected to a multiple system operator (MSO). One, identified as MSO, is a dummy variable, which is 1 if the system is connected to a multiple system operator and 0 if it is not. The other is the number of systems connected to the multiple system operator; in log form it is identified as LMS. These variables could measure cost differences between MSO and non-MSO systems, which could result from monopsony power or differences in the quality of service (or other nonprice dimensions of the product offering) not reflected in the other variables included in the regression. It may also reflect other behavioral differences between MSO and non-MSO systems. There are 2 dimensions of size reflected in our model. As noted above, these are both entered in our model in reciprocal form. One is the total number of channels, which is identified in reciprocal form as RTC. The other is the number of subscribers, for which we have used the number of subscribers in the system. It is identified in reciprocal form as RSS. There are 5 product mix variables in our model. One is the proportion of channels that are cable-only (i.e., non- broadcast), identified as PNB. The cable company usually has to pay a fee to the channel originator on a per- subscriber basis for these channels. Therefore, higher values of PNB should be associated with higher costs. Also, lower values of PNB may reflect a greater degree of competition with over-the-air broadcasting. The others are variables which relate to equipment or services for which subscribers are usually charged extra beyond the charge for basic service. As noted above, the revenues from these services are included in our dependent variable. Each of them is measured as the proportion of subscribers in the franchise that get these extra services, or, in the cases where subscribers may get more than 1, the ratio of the number of these services provided to the number of subscribers. These variables include the ratio of the number of additional outlets to the number of subscribers, identified as PAO; the proportion of subscribers to the second tier of channels beyond the basic tier, identified as PT2; the ratio of the number of tier subscription changes to the number of subscribers identified as PTC; and the ratio of the number of remotes rented to the number of subscribers, identified as PRM. The remaining variable in the model reflects other costs. It is the median income in the franchise area as determined from the 1990 Census. This variable, identified in log form as LIN, is included to reflect differences in wage rates of employees, and possibly other costs that might be correlated with wage rates. Regression Results Table A-1 shows the results of our regression estimates. The top part of the table shows the estimated coefficients and their standard errors. As noted above, the coefficients of the logarithmic variables can be interpreted directly as percentage impact estimates. However, the impacts of the proportion and dummy variables must be computed from the coefficients. The bottom part of the table shows those computed estimates of the impacts of changing the values of the variables from 0 to 1. As indicated above, the low penetration variable A1 has a coefficient that is not significantly different from zero, indicating that these franchises charge rates which are on average little different from the noncompetitive franchises. Its coefficient of -.010 implies that their rates are only about 1 percent lower on average than comparable noncompetitive franchises. OVL has a coefficient of -.174, which implies that the franchises facing competition throughout their service area have rates that are about 16 percent lower than comparable noncompetitive franchises. The municipal variable C has a coefficient of about -.231, which implies that the municipal sample has rates that are about 21 percent lower than comparable other competitive franchises and about 37 percent lower than comparable noncompetitive franchises. The regression estimate indicates that MSO franchises charge rates that are about 7 percent higher than non-MSO franchises, and that the franchises of MSOs with larger numbers of systems charge slightly higher rates than MSOs with fewer systems. A 100 percent increase in the number of systems in an MSO is associated with a 1 percent increase in rates. Other things being equal, franchises of large systems have lower rates than those of small systems. This probably reflects the spreading of fixed costs over more subscribers. Because of the nonlinear nature of the RSS variable, its impact is not as clear as the other variables. However, the primary impact of system size is for very small systems. The coefficient implies that a system with 810 subscribers charges an average rate that is only about 1 percent higher than that of the largest system. But that same system of 810 subscribers would have an average rate that is 36 percent lower than the smallest system in our sample, which has only 18 subscribers. Other things being equal, rates increase as the number of channels increase, but the relationship is nonlinear. The coefficient of RTC implies that an increase in the number of channels from 7 (the smallest in our sample) to 14 is associated with an 11 percent increase in rates, while an increase in the number of channels from 35 to 70 (the largest in our sample) is associated with a 2 percent increase in rates. The coefficient of PNB indicates that a franchise with no broadcast channels would have rates that are 29 percent higher than a franchise with all broadcast channels. A smaller and more realistic change of 10 percentage points (e.g., from 60 to 70 percent) is associated with rates that are less than 3 percent higher. Higher average rates are also associated with optional equipment and features. This is to be expected, because the charges for these options are included in these average rates (which are really average revenues per subscriber from regulated services). Franchises where there is an average of 1 additional outlet rented per subscriber have average rates that are 11 percent higher than those where no one rents additional outlets. Franchises where there is an average of 1 remote rented per subscriber have average rates that are about 19 percent higher than ones than those where no one rents remotes. Franchises where every subscriber subscribes to Tier 2 have average rates that are 6 percent higher than those where no one subscribes to Tier 2. Franchises where there is an average of 1 tier change per 10 subscribers have rates that are less than 4 percent higher than those where no one changes tiers. Higher labor costs, as estimated by the income variable are also associated with higher rates, with a 100 percent increase in labor cost being associated with rates that are about 7 percent higher. Other Variables Tried and Eliminated There were many other variables tested for inclusion in the model, either because we felt they might influence rates or because commenters argued that they should be included. In most cases they were eliminated because their coefficients were not statistically significantly different from zero at the 95 percent confidence level. However, as noted below, in a few cases the variables were excluded because of other problems. In most cases, the inclusion of these additional variables has little impact on the estimated coefficient of OVL. The few exceptions to this included values that were both lower and higher for that coefficient. Therefore, we believe that our estimate of the competitive differential is reasonably robust to alternate choices of variable inclusion. In our regressions we considered several variables that measured the length of time (in years) during which competition had been in place, as measured by the age of the newest headend among the competitive systems operating in a franchise area. Our expectation was that the competitive differential would diminish over time, because price wars are more common in the early stages of competition and collusion and non-price competition are more common in later stages of competition. Our best estimate involving this type of variable confirmed this, showing a competitive differential for the overbuilds that started at 24 percent for brand-new competition, then diminished to 17 percent after five years and to zero after fifteen years. The problem with this variable is that the oldest competitive systems were 26 years old, at which point the coefficient implies that overbuilds have rates that are 21 percent higher than non-competitive systems, which is implausible. Consequently, we rejected specific use of this variable in our final model because it leads to implausible estimates. We tried several other functional forms, however, which generally yielded the conclusion that the prices rise with the duration of competition. We did not pick any one variable or functional form as the right one. We tried an interaction between OVL and MSO. This variable did not have a significant coefficient. This would indicate that the competitive differential is approximately the same for MSOs and non-MSOs. We tried including separate dummy variables for the municipally owned systems and the privately owned franchises competing against the municipally owned systems. The coefficient for the privately owned franchises was more negative than that of the municipally owned systems, indicating that the private firms were charging rates that were lower than the model would predict for comparable municipally owned systems. However, the difference between the 2 coefficients was not statistically significant. Therefore, we decided to include just 1 variable, C, for the municipal sample. We tried various other functional forms for the subscriber variable. These included a linear version, a logarithmic version, and the reciprocal of the logarithm. The reciprocal of the number of subscribers gave the best fit. This form of the variable is compatible with the concept of fixed costs being spread over the number of subscribers. As an alternative to the use of system subscribers, we tried using franchise subscribers, again using various functional forms. This resulted in a slightly worse fit for the regression. When both system and franchise subscribers were included in the equation, only system subscribers had a significant coefficient. We also tried a subscriber variable analogous to OVL, which was equal to system subscribers if the same rates were charged to all franchises in a system, and franchise subscribers otherwise. This also yielded a worse fit than system subscribers. Therefore, we selected RSS as the best measure of the number of subscribers. As an alternative to RTC as the channel variable, we also tried the log of average total channels. This resulted in a worse fit. As alternatives to PNB, we also tried using the proportion of various other channel groupings, including satellite only, satellite and other, and non-local broadcast. These all resulted in worse fits, as did including the groupings of channels as totals instead of proportions. We also tried adding a dummy variable for six or more local broadcast signals. This variable did not have a significant coefficient. Various other product mix variables were also tried. These included installations per subscriber, disconnects per subscriber, reconnections per subscriber, churn per subscriber (the sum of the above three), tier 3 subscriptions per subscriber, converters rented per subscriber, and addressable converters per subscriber. These generally had insignificant coefficients. We also tried various other cost variables. These included percentage urban, line miles in the franchise area, line miles in the system, density (for the franchise area and the system), defined either as households passed per line mile or as subscribers per mile, percentage of plant below ground, percentage of fiber, and the age of the principal headend. In addition, we tried the amount of revenue per system subscriber from pay TV and advertising. None of these variables had coefficients that were statistically significant. It is not clear which of these cost variables should be reflected in rates if the firms are basing rates on marginal costs. We also tried various demographic variables derived from the 1990 decennial Census. These included percent of the population below the poverty line, percent urban, percent non-English speaking, percent white, percent of households containing a single person, percent owning their homes, and percent with children under the age of 18. These are all variables that are more likely to affect demand than supply. None of them had statistically significant coefficients. System Size and the Competitive Differential Several commenters pointed out that, using our original equation, the competitive differential was large and statistically significant for small cable systems but statistically insignificant for large cable systems. We were able to replicate these results with our cleaned-up data. We analyzed possible explanations for this finding and explored whether the apparent difference in the competitive differential by system size was either due to faulty measurement or to interactions between competition and other variables affecting revenue per subscriber. We expected that construction of a system overlap measure to improve the measurement of competition for large systems would eliminate the difference in the competitive differential by system size. As noted above, this variable has a larger and more significant coefficient and explains more of the variance than the franchise level overlap variable. But if a dummy variable for small (less than 5000 subscribers) competitive systems is added, the coefficient of the system overlap variable becomes insignificant. One possible explanation of the difference in the competitive differential with system size is that large systems facing competition are more sophisticated than small ones and have learned to collude more effectively with their rivals. If this explanation is correct, it suggests that the competitive differential for small systems is closer than that for the whole sample to the true effect of competition on price. Nevertheless, any partitioning of the data into size classes will be arbitrary, since we know nothing about the technology or economics of cable systems to suggest a rationale for the choice of size classes. Given the small size of the competitive sample, dividing it still further and attempting to draw inferences from a still smaller sample appears unwarranted. Thus we believe that any estimate used for rate regulation should be based on the entire sample. We tried interactions between OVL and system size, using various functional forms. In most cases, this interaction was not statistically significant or it resulted in a very high degree of multicollinearity with the OVL variable and yielded coefficients of unreasonable magnitudes. The one functional form that did seem to work was a linear one involving the product of OVL and system subscribers. The coefficient estimates using this form imply that rates for competitive franchises of larger systems rise to the level of non-competitive ones. But when extrapolated beyond the range of observed values for this interaction variable, the equation predicts implausibly high values for the dependent variable. In addition, in our sample of municipals and overbuilds, there is a high correlation between system subscribers and length of competition. We noted above that the competitive differential for overbuilds declines as the duration of competition increases, possibly because cable operators learn over time to engage in parallel pricing strategies. This leads us to conclude that if there is an interaction between competition and system size, it may take the form of franchises of large systems engaging in less vigorous price competition, and thus charging prices approximating those of the non-competitive sample. As a further attempt to isolate a possible interaction between competition and system size, we tried adding a variable which is equal to RSS for franchises that were included in the B and C samples and zero for those that were not. Unlike the system size interactions described above, this variable had a coefficient that did not have a very severe multicollearity problem. This coefficient, which can be treated as an estimate of the effect of system size on the competitive differential, was not significantly different from zero, further casting doubt on the interaction between competition and system size. Possible Simultaneous Equations Bias Whenever a supply or demand equation is estimated independently of the other, there is often a concern that what is being estimated is really a combination of both supply and demand factors. This is known as simultaneous equations bias. There are 2 ways this possible problem can be eliminated from the regression estimates. The easier, and the one used here, is to replace households subscribing (quantity demanded) in the equation with households passed (quantity supplied). The other would be to use a simultaneous equations estimation technique such as two- stage least squares. This technique allows for the interaction between supply and demand by introducing some extra variables that only affect demand. This would normally be a slightly preferable approach because the effect of the number of households on rates probably results primarily from the spreading of fixed costs, which is determined by the number of households actually subscribing, not the number of households passed. However, because of the way in which the number of subscribers enters the equation as a reciprocal, this approach yields an intractable functional form. Because the number of households passed is highly correlated with the number of subscribers, replacing RSS with RHP, the reciprocal of households passed in the system, will result in an estimate that is close to what could be achieved using a two-stage least squares approach. Table A-2 shows the estimates that result from this approach. A comparison of the regression results using this approach with the ordinary least squares results of Table A- 1 shows that, except for the coefficient of system size, the results are very similar. This indicates that there is no significant simultaneous equations bias in our original estimates. Use of the Benchmark Equation in the Going-Forward Methodology Our going-forward methodology permits cable operators to recover external costs, to adjust non-external costs for inflation, and to adjust non-external costs by a specified amount per channel when the total number of regulated channels changes. Non-external costs are also adjusted on a tier basis when cable operators restructure tiers. This subsection explains how these adjustments were calculated, based on our benchmark equation. When a cable operator increases the number of regulated channels offered, it may recover additional non-external costs. When a cable operator decreases the number of regulated channels offered, the amount of non-external costs that it may recover decreases. When a cable operator moves channels from one tier to another, the amount of non- external costs that it may recover per tier changes. In order to calculate these changes in permitted non-external costs, cable operators must utilize the table below. The entries in the table, which are based on our benchmark equation, indicate the amounts, in cents per channel per subscriber per month, by which cable operators will adjust their non-external costs. When the total number of regulated channels changes, the cable operator will calculate the average of the old and new total numbers of regulated channels and consult this table to find the applicable per channel adjustment factor. The total permitted adjustment is the product of the per channel adjustment factor and the change in total regulated channels. It is positive if the number of regulated channels has increased and negative if the total number of regulated channels has decreased. The change in allowed non-external costs per tier is equal to the change in the number of channels on the tier (positive for increases and negative for decreases) multiplied by the per channel adjustment factor. If the operator is merely restructuring tiers and there is no change in the total number of regulated channels, then the operator would find its total number of regulated channels in the table and note the corresponding per channel adjustment factor. Then, for tiers losing channels, the allowed non-external costs would decline by the adjustment factor multiplied by the number of channels removed from the tier. For tiers gaining channels, the allowed non-external costs would increase by the adjustment factor multiplied by the number of channels added to the tier. We have based the non-external cost recovery formula on our benchmark rate equation. The benchmark equation shows how average revenue per subscriber changes as the number of regulated channels offered changes. A change in the number of regulated channels will clearly affect non-external costs and may affect programming costs as well. The total number of regulated channels appears in two variables of the equation. One is the reciprocal of total regulated channels, which has a negative coefficient. The other is the ratio of nonbroadcast channels on regulated tiers to total regulated channels, which has a positive coefficient. Intuition suggests that the non-external costs of adding a channel are the same for broadcast and nonbroadcast channels. However, adding a nonbroadcast channel probably adds more to programming cost than does adding a broadcast channel. (Must carry stations, for example, have no programming cost associated with them.) This leads us to conclude that the ratio of nonbroadcast channels variable primarily reflects changes in programming costs and the total channels variable primarily reflects changes in non- external costs. Suppose, however, that a cable system added a single broadcast channel with no associated change in programming costs. The value of the ratio of nonbroadcast channels variable would fall, reducing its contribution to average revenue per subscriber, even though programming costs did not change. The value of the reciprocal of total channels variable would also fall. Given the negative sign on its coefficient, this increases its contribution to average revenue per subscriber. That increase must include not only increased non-external costs but some component to counterbalance the change in the ratio of nonbroadcast channels variable. These considerations lead us to vary only the value of the reciprocal of total channels variable when we calculate the impact of channel changes, setting the variables in the equation other than the reciprocal of average total channels variable equal to their means for our noncompetitive sample of cable systems. Because the reciprocal of total channels variable likely includes some of the programming cost effect when nonbroadcast channels are added, we considered "backing out" a programming cost component from the permitted cost increases generated by our benchmark formula. However, it is impossible to determine how much of programming costs might be captured by this variable. Moreover, we have no good data on programming costs as a share of total costs. These considerations, plus our desire to avoid errors that might generate excessively low per channel adjustment factors, led us to refrain from reducing the channel adjustment factors generated by the benchmark equation in order to remove a programming cost component from them. To calculate the per channel adjustment factors in the table, we substituted mean values from our non-competitive sample into the benchmark equation for all variables other than the reciprocal of total regulated channels. Because the competitive dummy variables take on the value of zero in the noncompetitive sample, this means that we did not reduce the predicted average revenue per subscriber by our 17 percent competitive differential. The distinction is important because our procedure generates percentage changes in average revenue per subscriber as total regulated channels change. The magnitude of the change depends on the base value of revenue per subscriber. This magnitude is higher than it would be if we had chosen to deduct the competitive differential first. Our procedure here guards against excessively low channel adjustment factors and may be rationalized by noting that our benchmark equation generates changes in revenue per subscriber that were sufficient to make noncompetitive cable systems willing to increase the number of channels offered. Using the relation between average revenue per subscriber and average total channels described in the previous paragraph, we calculated the change in average revenue per subscriber for a large number of one and two channel increments and associated those changes with the midpoint of the increment. For example, the change from 10 to 11 channels is associated with 10.5 channels. The change from 10 to 12 channels (divided by two) is associated with 11 channels, etc. The cable systems in our noncompetitive sample had channel capacities ranging from seven to 70. We therefore performed this exercise over a range (six to 71 channels) large enough to generate adjustment factors for seven to 70 channels. The portions of that range with identical cost changes are grouped together in the table. The calculated cost changes are largest for small numbers of channels, reflecting the fact that our benchmark curve drops steeply at those channel levels and then flattens out substantially at higher channel levels. Low channel capacity systems are likely to have different, perhaps significantly different, characteristics than the average for the entire sample. Using sample averages could inflate the adjustment factors for those channel levels. Therefore, we made a separate calculation for the seven to 20 channels region, setting the variables in the equation other than the reciprocal of average total channels equal to their means for the subsample of noncompetitive systems with 20 or fewer channels. The figures in the table for seven to 20 channels are calculated in this fashion. For the range from 21.5 to 70 channels, the figures in the table are calculated using mean values for the entire noncompetitive sample as described above. We note that the number of subscribers to systems in the seven to 20 channels range is relatively small and that most of these systems are likely to be small systems, for which we plan to do more detailed cost studies. Hence, it is likely that we will have more detailed information on costs of systems in this channel range before many of those systems would have occasion to use this per channel adjustment table. Table A-1 Regression Results Variable Coefficient Low Penetration Dummy (A1) -.010 (.016) Overbuild Overlap (OVL) -.174 (.033) Municipal Dummy (C) -.231 (.050) MSO Dummy (MSO) .070 (.030) Log of MSO Size (LMS) .0097 (.0042) Reciprocal of System Subscribers (RSS) 8.14 (1.60) Reciprocal of Average Total Channels (RTC) -1.45 (.61) Proportion of Non-Broadcast Channels (PNB) .253 (.079) Proportion of Additional Outlets (PAO) .103 (.026) Proportion of Remotes (PRM) .172 (.031) Proportion of Tier 2 Subscribers (PT2) .057 (.019) Proportion of Tier Changes (PTC) .353 (.111) Log of Median Income (LIN) .069 (.021) Intercept 2.04 (.23) R Square .469 A1 impact -.010 OVL impact -.160 C impact -.206 OVL + C impact -.366 MSO impact .073 PNB impact .288 PAO impact .109 PRM impact .187 PT2 impact .058 PTC impact .424 Note: Numbers in parentheses are standard errors. Table A-2 Estimates Eliminating Possible Simultaneous Equations Bias Variable Coefficient Low Penetration Dummy (A1) .009 (.016) Overbuild Overlap (OVL) -.174 (.033) Municipal Dummy (C) -.217 (.050) MSO Dummy (MSO) .075 (.030) Log of MSO Size (LMS) .0108 (.0042) Reciprocal of System Households Passed (RHP) 21.42 (5.19) Reciprocal of Average Total Channels (RTC) -1.53 (.65) Proportion of Non-Broadcast Channels (PNB) .243 (.079) Proportion of Additional Outlets (PAO) .109 (.027) Proportion of Remotes (PRM) .171 (.032) Proportion of Tier 2 Subscribers (PT2) .054 (.019) Proportion of Tier Changes (PTC) .362 (.112) Log of Median Income (LIN) .070 (.021) Intercept 2.03 (.23) R Square .458 A1 impact .009 OVL impact -.159 C impact -.195 OVL + C impact -.354 MSO impact .078 PNB impact .275 PAO impact .115 PRM impact .186 PT2 impact .056 PTC impact .436 Note: Numbers in parentheses are standard errors. Table A-3 Adjustment Factors for Calculating Changes in Allowed Network Costs Average Number of Channels Per-Channel Adjustment Factor Average Number of Channels Per-Channel Adjustment Factor 7 $0.52 14 0.14 7.5 0.45 14.5 0.13 8 0.40 15-15.5 0.12 8.5 0.36 16 0.11 9 0.33 16.5-17 0.10 9.5 0.29 17.5-18 0.09 10 0.27 18.5-19 0.08 10.5 0.24 19.5-21.5 0.07 11 0.22 22-23.5 0.06 11.5 0.20 24-26 0.05 12 0.19 26.5-29.5 0.04 12.5 0.17 30-35.5 0.03 13 0.16 36-46 0.02 13.5 0.15 46.5 and up 0.01 February 22, 1994 SEPARATE STATEMENT OF COMMISSIONER ANDREW C. BARRETT RE: Implementation of the Cable Television Consumer Protection and Competition Act of 1992 -- Rate Regulation (Fourth Order on Reconsideration, Fourth Report and Order) With today's actions, the Commission revises its cable rate regulations by modifying the benchmark methodology, which serves as the primary approach for regulating cable service rates. In a separate rulemaking, the Commission establishes requirements to govern cost-of-service showings to justify rates above the levels determined by the benchmark approach. The Commission's decision affirms a benchmark methodology and establishes a new competitive differential at 17% relative to September 1992 rate levels to guide rate reductions. Accordingly, the revised rules will require systems to reduce rates by 17% from their September 1992 level, or to the new benchmark, whichever is less. Once systems make their necessary reductions to comply with the new benchmark mechanism, they are permitted to add external costs and to apply a "going forward" adjustments for additional channels or system upgrades. Systems that have reduced rates by 17% (i.e., a prior 10% adjustment under the old benchmark and an additional 7% under the new benchmark), also may make adjustments for inflation. In addition, these revised rules will initiate cost studies to verify cost differences among cable operators in comparison to the competitive differential. Systems that are required to reduce their rate by an amount less than the full 17% competitive differential -- as well as systems with rates below the new benchmark level that are not required to make any immediate reductions -- will be required to engage in future rate actions in accord with the results of the cost studies. As further elements of the cable rate regulation package, the Commission establishes (1) a mechanism to allow "going forward" adjustments for additional channels and system upgrades, and (2) a standard for targeted rate relief, as well as provisions for administrative relief, to small operators. During this proceeding, I have consistently stated that the Commission must implement rate regulations in an orderly and effective manner in order to maintain the integrity of our regulatory process, to avoid creating potential unintended consequences, and to minimize false expectations among the consumer public. I have also stated that the Commission's rate regulation mechanisms must (1) incorporate measures of flexibility in order to balance the concerns of the industry, consumers, and franchising authorities, and (2) minimize the uncertainty that has resulted from the cable rate regulation proceeding so that consumers and the industry may develop realistic expectations and business plans, respectively. I write separately today in order to emphasize that my decision to support this rate regulation package is based on the measure of flexibility built into a benchmark system of regulation, including several of the "going forward" and "cost-of-service" components. Given the lack of complete information on pricing and costs, and our relatively limited sample for competitive and noncompetitive pricing behavior, I believe that a revised benchmark approach exercises the necessary caution in recognizing the variety of cost structures and pricing practices throughout the cable industry. During the reconsideration process, the Commission has revised the benchmarks by correcting the data on competitive and noncompetitive systems as well as refining the statistical procedure for estimating the benchmarks. Therefore, I believe that the benchmark information, although arguably subject to certain shortcomings detailed in this proceeding's record, now forms a better foundation for other components of the rate regulation package, especially the "going forward" allowances for channel additions and upgrades. Next, I believe that the other components in this rate regulation package -- including the "going forward" methodology, the presumptions established to guide decisions regarding "a la carte" practices, and the provisions for a measure of small system relief -- will provide necessary flexibility to allow operators to begin to develop future business plans and to add new programming services. With respect to the "going forward" mechanism, I believe that the allowance for actual programming costs may help to avoid unintended consequences for program services as a result of the revised rate regulations. The opportunity for a streamlined cost-of- service showing also will allow operators to account for new services through upgrades of their systems. In addition, I support today's effort to distinguish legitimate "a la carte" marketing practices for programming services from those practices that could constitute evasions of the Commission's rate regulations. In this regard, I believe that the presumptions regarding "a la carte" practices will enable the Commission to identify legitimate package offerings that increase realistic consumer choices and provide for a reasonable number of programming services at favorable rates. Finally, the Commission has provided for a measure of rate relief for small operators, which will allow certain small operators to make external cost and "going forward" adjustments to rates regardless of where the rates of these systems fall relative to the benchmark. Nonetheless, I remain concerned that some small operators may find that further relief is necessary in order to avoid particular hardship, and I emphasize that the provision for additional hardship relief to certain small operators, as well as the streamlined cost-of-service mechanism, will become important recourse for small operators in such dire situations. The revised rules also will initiate a cost study to verify cost differences among cable operators in comparison to the competitive differential. I believe that this study will provide important information to guide the Commission's analysis of the differences among competitive and noncompetitive operators, as well as the operating distinctions that may exist among small, medium and large operators. I also believe that this detailed cost information will enable the Commission to evaluate the validity of many policy assumptions that have guided our efforts in this proceeding, and therefore, will help to identify whether further adjustments are necessary to these refined rate regulations. As a result, I believe that it is appropriate to postpone rate actions as applied to certain systems, especially to small systems and those systems with rates below the new benchmark level, and to base future rate actions for those systems upon the differential, if any, identified by results of the cost study. I especially am interested in the cable industry's full participation in this cost study in order to resolve a notable void in this proceeding's record. As a consequence, I believe that these studies must be completed as soon as possible before the end of 1994 in order to promote the certainty that will enable all operators to develop future business plans. Based upon my own analysis, I believe that the new competitive differential of 17% as compared to the September 1992 rate levels represents the highest point of what I consider to be an acceptable range for this policy determination. I have previously asked questions regarding the proper procedure for calculating the differential between competitive and noncompetitive rates, especially concerning the effect of the statistical treatment for low penetration and municipal systems on the differential. I am aware that the Commission's revised data and statistical procedures provide analytical support for a 17% differential by focusing primarily on the differences between noncompetitive systems and overbuild systems, while retaining a measured consideration of the low penetration and municipal systems in the competitive sample. Nonetheless, I remain concerned that a more cautious approach for developing a competitive differential would reflect the limited confidence that results from a relatively small sample size and a lack of cost data. Furthermore, I consistently have emphasized the need to consider the effect of the cable rate regulations on industry investment. The freeze on cable revenues and the implementation of the benchmark mechanism have had a negative impact on the cable industry's revenues and kept rates from rising. For example, a recent study states that these actions "have already precipitated more than an estimated $2 billion direct loss of revenues and cash flow," while also citing "the complete foreclosure of growth avenues for cable TV programmers." Therefore, I am concerned about the potential effects of the 17% competitive differential relative to the September 1992 rate level. However, I also believe that the entire cable rate regulation package, including our cost-of-service and "going forward" options, incorporates important elements of flexibility that will allow operators to adjust to the 17% differential, where necessary. For example, all systems are permitted to adjust rates for external costs and "going forward" factors. Systems above the benchmark, which are required to reduce their rates by the full competitive differential, are permitted to make adjustments for inflation as accrued between September 1992 and September 1993. Additionally, I believe that the benefits of the"going forward" mechanism for many operators will occur through the streamlined cost-of-service process, which will be subject to further comments and refinements. Given that this process will affect the incentives for operators to invest toward future system developments and the carriage of new programming services, I am concerned that this streamlined cost showing serve as an important bridge between the benchmark mechanism and the requirements for a full cost-of service showing. The streamlined cost-of- service process will play a critical role within the rate regulation framework, especially through provisions for an incentive-based plan for upgrades and the opportunities to demonstrate separate allocations for improvements to existing regulated services. With respect to the cost-of-service proceeding, I support various aspects of the Order that grant flexibility to operators with unique cost-based circumstances that justify rates above the new permitted benchmark level. Furthermore, the cost-of-service process includes a rate-of-return factor of 11.25% that is reasonable as compared to other regulated industries, especially after tax considerations are included. I believe that necessary flexibility in the cost-of-service process also occurs through the cost allocation mechanism, the procedure for determining the portion of excess acquisition costs that operators may recover, as well as the provisions for treatment of Subchapter S corporations. As a result, I believe that the range of factors considered in the cost-of-service process, including the option for hardship showing, will begin to mitigate some of the consequences for cable operators who may endure the most significant changes as a result of the new 17% competitive differential. Finally, I believe that this decision must be viewed in light of the overall package of elements that affect the rate calculations as well as the rate adjustments and cost showings that will be allowed. Therefore, I encourage the industry to await the release of all final orders before assessing the effects of these decisions on their particular markets. In the end, my goal is to ensure that our decisions in this area are balanced and will permit continued investment to enhance services to the public. These rulemakings on cable rate regulation have involved extremely complex analysis, and I acknowledge the outstanding dedication shown by our Commission staff, my colleagues, and their respective staffs.