Skip to main content

Declaration Of Bridger M. Mitchell, CRA International

This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.




August 8, 2007


  1. Qualifications and Summary
  2. Market Definition Issues
    1. Product Market
      1. Channel Termination Services and Channel Mileage Services are Distinct Product Markets
      2. Channel Termination Services and Channel Mileage Services Should be Treated as if DS1 and DS3 Circuits are in Distinct Product Markets

    2. Geographic Market
      1. The Wire Center is a Useful Approximation of the Relevant Geographic Market

  3. There are Significant Barriers to the Emergence of Competition in the Provision of Special Access Services
  4. As a Result of Barriers to Entry, There is Little Competition in the Supply of Special Access Services
  5. Integrated BOCs have Incentives to Raise Wireless Rivals' Costs of Special Access
  6. The Exercise of Market Power: the Pricing Evidence
    1. Comparing Rates Under Pricing Flexibility and Those Under Price Caps
    2. Contract Pricing Plans Enable BOCs to Extend Market Power
    3. Summary

  7. Conclusion

Attachment A: Curriculum Vitae

Before the


Washington, D.C. 20554

In the Matter of

Special Access Rates for Price Cap Local
Exchange Carriers

AT&T Corp. Petition for Rulemaking to Reform
Regulation of Incumbent Local Exchange Carrier   
Rates for Interstate Special Access Services

WC Docket No. 05-25


Declaration of

Bridger M. Mitchell

August 8, 2007

  1. Qualifications and Summary
  1. My name is Bridger M. Mitchell and I am a Vice President at CRA International, an economics, finance, and business consulting firm with 23 offices across North America, Europe, Asia Pacific, and the Middle East. I am an expert in competition and pricing in the telecommunications industry and have provided expert testimony, litigation support, and economic consulting services to numerous business and government clients. My research on major regulatory issues encompasses the theory and practice of telecommunications pricing, competition, and equal access in local telephone markets, interconnection in telecommunications networks, international telephone rates, and broadcasting and cable television. I have developed pioneering models of the cost structure of a cable television firm and the incremental costs of local telephone networks. I previously taught economics at Stanford University and UCLA and was a senior economist at The RAND Corporation. I have a Ph.D. in Economics from the Massachusetts Institute of Technology. My CV is included as Attachment A to this declaration.
  2. I have been asked by Sprint Nextel to review and update the declaration that John Woodbury and I filed in this proceeding in 2005 on behalf of Nextel.(1) In that declaration we reviewed comments filed in this proceeding for the purpose of evaluating the effects of the Commission's triggers for granting the Bell Operating Companies (BOCs) flexibility in the pricing of special access services.
  3. I update our 2005 analysis to assess "whether actual marketplace developments support the predictive judgments that underlie the special access pricing flexibility rules."(2) Based upon developments since that original analysis, I conclude that, in fact, in those Metropolitan Statistical Areas (MSAs) in which the BOCs received pricing flexibility, competition has not emerged or had the salutary effects anticipated by the Commission.
  4. The special access services of principal concern here are channel termination (CT) and channel mileage (CM). CT provides the link between the end customer location and the BOC central office. CTs have the same basic economic characteristics as a local loop, requiring high fixed and sunk costs to serve an individual customer location. CM provides interoffice transport between two BOC central offices.(3) For Sprint Nextel and other Commercial Mobile Radio Service (CMRS) customers, CTs and CMs are required to connect individual cell sites located throughout an MSA to the wireless customer's mobile switching center.
  5. In the discussion that follows, I review the analytic steps required to evaluate whether or not competitive conditions in Phase II pricing-flexibility MSAs have emerged and have successfully constrained the BOC rates for these special access services. The first step is to define the relevant markets, both product and geographic. Among other issues, I explain why the geographic and product market definitions used by the current FCC pricing flexibility rules are flawed. Instead, for purposes of developing criteria that indicate whether or not competition is supportable, I conclude that the product market should distinguish between the different CTs and CMs (and their capacities), and that the geographic market should be much narrower than the MSA.
  6. Having defined the relevant markets, I then briefly consider the barriers to emerging competition in these markets and conclude that such barriers can be substantial, particularly on routes with thin traffic demands.
  7. The existence of substantial entry barriers leads to the inference that on many routes, the degree of competition in special access is likely to be quite limited. Sprint Nextel's own experience in the purchase of special access services (as well as the experience of other parties) provides empirical evidence of the lack of competition.
  8. The combination of substantial entry barriers and the general lack of special access competition leads to the prediction that the BOCs can maintain or increase prices of special access to supra-competitive levels in the absence of any regulatory constraint. The evidence that has been provided to the Commission clearly supported such an inference in 2005 and more recent information reconfirms this inference. In those MSAs where the BOCs have been granted pricing flexibility, special access rates have not fallen; in many cases they have increased. Moreover, the gap between the price-capped rates and the unconstrained BOC rates has typically increased over time, and the price-flexibility rates have remained substantially above the incremental costs of at least one set of comparable offerings: high-capacity unbundled network element (UNE) loops and transport.(4)
  9. Thus, the evidence indicates that purchasers of special access services, like Sprint Nextel and other CMRS carriers, have been paying prices significantly higher than the levels an effectively competitive marketplace likely would have produced and higher than would have been the case had the FCC's prior price-cap regime remained in force in those MSAs where the BOCs have obtained pricing flexibility. In short, the pricing flexibility triggers did not result in lower special access rates for purchasers.
  10. I conclude that the entry barriers to the provision of special access services are substantial and the existence of those barriers has prevented special access competition from emerging throughout the MSA. And the effect of allowing the BOCs flexibility in pricing special access in areas where competition is non-existent or nearly so has elevated special access rates further above competitive levels. As a result, consumers have been harmed through these higher prices. The Commission should revise its pricing flexibility rules so that they properly reflect product markets distinguished by type of special access service and capacity; geographic markets that are route-specific; and criteria that reflect the extent of sunk costs and the attainment of scale economies required for competitive entry.
  11. The next section addresses questions of market definition. Subsequent sections examine the extent of entry barriers to new competition; the extent to which competition in the provision of special access services has in fact emerged; the incentives of integrated BOCs to raise wireless, wireline and broadband rivals' costs; and the pricing effects of the Commission's pricing flexibility rules.

  1. Market Definition Issues
  1. The FCC established appropriate market definitions in the TRRO(5) for unbundled network elements, and should use those same definitions in assessing competition for special access services. This section summarizes the special access product and geographic markets that are relevant to CMRS carriers and other purchasers of special access.

  1. Product Market
    1. Channel Termination Services and Channel Mileage Services are Distinct Product Markets
  1. The definition of a market requires a demand-side analysis, querying whether a hypothetical monopolist of special access services can profitably raise price, holding other things equal, including the number and extent of service supplied by all firms in a market.(6)
  2. A CT is required to provide a circuit from a customer's premises (e.g., a Sprint Nextel cell site) to the serving wire center. CM service is required to extend that circuit from the serving wire center to a more distant wire center or central office, from which it can then be connected to a customer's point of interconnection (e.g., Sprint Nextel's mobile switching center). For the customer, CTs and CMs are not substitutes. A small increase in the price of CMs by a hypothetical monopolist will not lead customers to increase their purchases of CTs so as to render that price unprofitable. Similarly, a small increase in the price of CTs will not lead customers to increase their purchases of CMs. This is so because CTs connecting cell sites to serving wire centers cannot substitute for CMs connecting BOC central offices, and vice versa. CT and CM services are therefore distinct product markets. Indeed, CT and CM services are more likely complements than substitutes.

    1. Channel Termination Services and Channel Mileage Services Should be Treated as if DS1 and DS3 Circuits are in Distinct Product Markets
  1. Special access circuits are also distinguishable by differences in capacity (e.g., DS1 vs. DS3).(7) The competitive availability of DS1 service and DS3 service will likely differ markedly by geographic area. Distinguishing between DS1 service and DS3 service is critical to evaluating the extent to which competition may or could exist in different geographic areas, and thus these distinctions are critical in adopting criteria that will more accurately indicate the availability of likely competition in CTs and CMs.
  2. As the Commission already found in the context of the TRRO, competitive supply of stand-alone DS1 CTs is uneconomic.(8) Instead, when competitive carriers provide DS1 CTs they do so by channelizing their own DS3 facilities (or by leasing DS1 channelized services from a DS3 provider). When another customer in the same building is already served by competitive fiber, the incremental costs of channelizing a DS3 facility to provide DS1 loops are minimal.(9) By contrast, deployment of a fiber lateral to a building in order to serve fewer than seven DS1s would likely be uneconomic.(10)
  3. As a result, the availability of competitive DS1 loops throughout a particular geographic area will be highly correlated with the availability of competitive DS3 loops and the level of demand for those DS3 services by carriers such as Sprint Nextel. The availability of competitive DS3 CTs in individual buildings will depend on the demand for DS3 and higher-capacity services in those buildings and the area served by the relevant serving wire center. A large number of business lines in a wire center service area increases the potential revenue opportunities for DS3 service within the individual buildings located in that area.
  4. In areas outside the central business district, the demand for special access services is likely to be for DS1, but not DS3, levels of capacity. The Commission itself concluded in the TRRO that "the majority of small and medium-sized business customers occupy single tenant commercial buildings and that the building of laterals for DS1 services requires many customers at a single location to justify their costs."(11) In these areas, there will likely be relatively few business lines and only limited fiber collocation by competitive carriers. Consequently, there is a low likelihood of competitive DS3 facilities at a large proportion of the customer locations. An increase in price by a hypothetical monopolist of DS1 CTs is therefore unlikely to be defeated because it is unlikely that competitive DS3s will be available at those locations. Instead, the BOC is likely to be the dominant provider in markets with limited demand for high-capacity circuits because it can provide stand-alone DS1 CTs over its legacy copper facilities.
  5. Now consider a price increase by a hypothetical monopolist of DS3 CTs. Such an increase could be rendered unprofitable only if there were providers of channelized DS1 services already serving the building served by the DS3 providers and if those providers of DS1 services reconfigured the electronics on their DS3 facilities to provide DS3 services. But, as a practical matter, this means focusing on the availability of DS3 facilities only, as those DS1 providers would exist only if there were sufficient demand to warrant the construction of competitive DS3 facilities in the first place. Many markets outside the central business district lack the revenue opportunities that would lead competitive local exchange carriers (CLECs) or other providers to enter and sink the costs of DS3 facilities, resulting in the incumbent local exchange carrier (ILEC) being the dominant provider of DS3 CTs over its existing facilities.
  6. The same kinds of capacity distinctions should be made with respect to CM markets. On many routes outside the central business district, lesser demands for transport are more likely to require only DS1 rather than DS3 service. In less densely populated areas, the transport lengths are likely to be longer than in more densely populated areas, increasing the costs of deploying a DS1 circuit. In these markets, the likelihood that there will competitive alternatives to the BOCs' transport facilities may be quite remote. Were DS3 transport available, it could be channelized to offer DS1 transport (so that DS1s and DS3s might be substitutes from a demand/functional equivalence perspective). However, there will not likely be any DS3 transport providers in these areas, including wholesalers, to constrain a hypothetical monopolist of DS1 transport services. In markets characterized by limited demands for high-capacity transport, competitive DS1 transport may not be profitable and the ILEC is likely to be the dominant provider of DS1 services.
  7. On routes with greater demands for high-capacity transport, both DS3 transport services and DS1 transport (channelized on a DS3) may be available. In these markets, an increase in the price of DS3 service by a hypothetical DS3 monopolist could result in providers of channelized DS1s using their DS3 capacity to compete with the hypothetical DS3 monopolist; similarly, an increase in the price of DS1 service by a hypothetical DS1 service monopolist could lead DS3 providers to channelize their capacity to provide DS1 service. However, in both cases, it is the availability of sufficient demand to support sinking the costs into DS3 transport capacity that allows such competition. Thus focusing on the likelihood of actual or emerging DS3 transport competition is appropriate in such markets. In markets with insufficient demand to support DS3 competition, the ILEC will be the sole provider of DS3 service.
  8. These observations echo the conclusions of the Commission in the TRRO with respect to high-capacity loops and transport, conclusions that are equally relevant for CTs and CMs. With regard to CTs, the practical question is whether it is likely that other competitive carriers have already deployed, or will likely deploy, DS3 facilities throughout the wire center serving area, thereby making (channelized) DS1-level supply of those deployed facilities potentially viable.(12) As the Commission also noted,(13) where demand for high-capacity loops exists only at the DS1 level of service, there is insufficient traffic for competitive suppliers to enter with DS3 facilities and supply DS1 loops, and the analogous conclusion can be reached for channelized DS1 CTs.(14)
  9. The Commission reached similar conclusions with respect to interoffice transport that should pertain to CMs as well. For example, in the TRRO, the Commission concluded that "a carrier requiring only DS1-capacity transport between two points typically does not have a large enough presence along a route (generally loop traffic at a central office) to justify incurring the fixed and sunk costs of self-providing just that DS1 circuit."(15) While the Commission was referring to competitive carriers, the position of carriers like Sprint Nextel with respect to self-provisioning is similar to that of a competitive carrier.(16)
  10. Thus, in evaluating the extent of competition in the provision of DS3 service, one only need assess existing DS3 capacity. In markets where the demand for high-capacity services is limited and the competitive deployment of DS3 facilities is unlikely, there will be no competitive DS1 constraint on DS3 prices. In such markets, the BOC may be the only, or dominant, provider of DS3 services.
  11. Where the demand for high-capacity services is limited, the rates for DS1-capacity CTs and transport will not be constrained by competitive DS3s because competitive DS3 capacity is not likely to be deployed. In these markets, the BOC is likely to be the dominant provider of DS1 and DS3 services. Thus, as a practical matter, in determining appropriate criteria for supportable competition for CTs and CMs, the Commission should distinguish between the viability of competition for DS1 service and for DS3 service.

  1. Geographic Market
    1. The Wire Center is a Useful Approximation of the Relevant Geographic Market
  1. In revisiting the standards for requiring access by competitive carriers to unbundled network elements, the Commission posed the demand-side question with respect to the geographic scope of the market and (referring back to earlier decisions) concluded that customers' demand was route-specific.(17) For example, a customer seeking transport between points A and B would not find transport between A and C a good substitute. Similarly, the FCC found that "a loop serves a specific location and cannot economically be transferred to serve another customer location."(18) Thus, the FCC determined that competition at the level of the wire center would be a reasonable proxy for competition for customers seeking services originating, terminating or going through that geographic point.
  2. The Commission should use that same analysis here to conclude that using the MSA as a geographic market definition is inappropriate for special access as well. Indeed, in the TRRO, the Commission soundly rejected the notion that regions at least as large as MSAs are the appropriate geographic market — finding that "an MSA-wide approach would require lumping together areas in which the prospects for competitive entry are widely disparate."(19)
  3. The experience of CMRS carriers in purchasing special access services illustrates the limitations of using the MSA as the relevant market. CMRS carriers don't just offer service in the central business district — they offer service with a national footprint that includes areas within an MSA with only limited demands for high-capacity circuits. It remains as true today as it was in 2005 that CMRS carriers such as Sprint Nextel require special access services for circuits that connect their cell sites to mobile switching centers, and those cell sites are spread widely throughout an MSA.

  1. There are Significant Barriers to the Emergence of Competition in the Provision of Special Access Services
  1. The ability of BOCs to raise and maintain special access prices above competitive levels depends critically on the profitability of entry by new competitors and the profitability of expansion by existing competitors. However, sunk costs and the need to achieve economies of scale and scope constitute significant barriers to profitable entry and expansion by competitors and thus to effective competition in most of the markets for DS1 and DS3 special access services. Sprint Nextel's experience in purchasing special access services reflects this lack of competition.
  2. The Commission, in its extensive examination of competition in loops and interoffice transport in its TRRO, has already found that there are significant barriers to competitive supply of loop and transport services (analogous to CTs and CMs, respectively). The FCC found that loop investments require large sunk costs and therefore pose substantial barriers to competition for high-capacity loops.(20) This conclusion is consistent with a pre-merger AT&T analysis that quantified the importance of sunk costs and found that competitive carriers must have at least three DS3s of demand at the potential location to justify deploying CTs.(21)
  3. That analysis found that the lion's share of the cost of transmission facilities required by a competitive carrier is sunk and would be stranded and lost if it is unable to offer service profitably.(22) And pre-merger AT&T's analysis found that deployment of transport facilities to a particular point of aggregation is only economic when there are at least 18 DS3s of traffic available.(23)
  4. Moreover, as discussed previously, CLEC investments in the most profitable areas of an MSA do not enable competitors to provide CT or CM service in many or most of the other areas within the MSA. Consequently, in the great majority of the areas, sunk costs constitute a sufficient barrier to entry to make the BOC an effective natural monopoly in CT service. Moreover, in their analysis Drs. Ordover and Willig found that sunk costs and limited traffic concentration also provide the conditions for what is effectively natural monopoly in many interoffice CM markets.(24)
  5. In addition to significant sunk costs, special access services are supplied under conditions of economies of scale and scope. Fixed costs of trenching and laying cable, combined with lower unit costs of both higher-capacity fiber and electronics, provide the supplier who achieves greater aggregate demand in a geographic market with a significant cost advantage over competitors with lesser demand. BOCs enjoy economies of scale by aggregating the demands of customers located along a route. In addition, they achieve economies of scope on high-capacity transport facilities by combining special access traffic with traffic from other BOC-supplied services (local telephone, long-distance, data and IP-based services, etc.).(25)
  6. Since the filing of comments in 2005, the mergers of MCI and Verizon, SBC and AT&T, and AT&T and BellSouth have increased the economies of scale and scope that are achievable by BOCs and provided them an increased competitive advantage over other suppliers of special access services. All of these incumbents have cited cost savings as benefits of their respective combinations. Verizon expects benefits of its merger with MCI to make it a more efficient competitor and to realize financial benefits from eliminating overlapping network facilities.(26) SBC expects synergies of merging with AT&T to reduce costs of network operations as facilities and operations are consolidated.(27) And AT&T cited its ability to integrate wireless and wireline IP networks with BellSouth to realize substantial cost savings.(28)

  1. As a Result of Barriers to Entry, There is Little Competition in the Supply of Special Access Services.
  1. The barriers to entry by competitive suppliers of special access services reviewed above have resulted in only limited competition in the supply of special access services. In this section, I test the BOCs' claims that competition is extensive(29) against the evidence of the actual extent of competition.
  2. Sprint Nextel's nationwide CMRS network encompasses virtually every MSA. Overall, Sprint Nextel purchased 97 percent of DS1 circuits for both its wireless and wireline networks from ILECs in 2006.(30) For DS3 circuits, the corresponding percentage is 88.4 percent.(31) Thus, most of the markets in which Sprint Nextel requires special access services in fact have only very limited competitive alternatives.
  3. In 2004 Nextel issued a request for information (RFI) for the provision of high-capacity CTs and CMs in the New York City metropolitan area. The responses provide further evidence of the lack of competitive alternatives. Nextel had over 1,500 cell sites that were the focus of the RFI. Even though New York City is arguably the most fertile ground for the development of special access competition, CLECs responded by offering to provide special access for only 43 of those 1,500 cell sites.(32)
  4. Sprint Nextel's experience is not atypical. In a 2005 filing, another CMRS carrier, T-Mobile, similarly reported that it has depended on ILECs for over 96% of its CT links.(33) Even pre-merger AT&T, at that time one of the largest CLEC suppliers of special access, obtained 93% of its DS1-level transport from incumbent carriers, according to a 2003 Declaration submitted by pre-merger AT&T.(34)
  5. The record in this proceeding contains substantial reliable information regarding the lack of competition in special access. The most extensive evidence has been developed by ETI. In a 2003 declaration on behalf of pre-merger AT&T, Lee Selwyn determined that in MSAs that have some Phase II pricing flexibility, pre-merger AT&T relied on ILECs for special access services in about 94% of the buildings served by pre-merger AT&T.(35) In MSAs where there is no pricing flexibility, that percentage rose to 97%.(36) Thus, pre-merger AT&T's reliance on CLECs was nearly the same in both categories of MSAs, suggesting that the availability of competitive alternatives to the BOC is equally limited in both types of MSAs. Even in what is arguably the most promising metropolitan area for CLEC competition, the New York MSA, no non-ILEC facilities were available at nearly 86% of the business locations served by pre-merger AT&T.(37) And in New York City, only 900 of 220,000 buildings (mixed use, commercial, industrial, and public) were served by CLEC fiber at the time of the Selwyn Declaration in 2003.(38) Indeed, as the subsequent ETI study highlighted, non-ILECs provided facilities-based service to no more than 1% of the commercial buildings in the United States.(39)
  6. The competitive landscape has not improved measurably since that time. In January, 2007 Sprint Nextel queried 77 alternative access providers asking whether they could provide special access services to any of more than 52,000 Sprint Nextel cell sites nationwide. Just 16 vendors had such capability at any of those Sprint Nextel cell sites and those competitive vendors were equipped to serve only some 1% of the cell sites with on-net facilities.(40) The responses confirm the lack of competitive alternatives in most areas of Sprint Nextel's network.
  7. Even if transport services were supplied in competitive markets (conditions that may occur more frequently than for CTs), the BOCs may be able to capture the monopoly profits from the sale of price-capped CTs. Specifically, wireless carriers typically (but not always) purchase CTs and CMs for the same circuit from a single vendor and at separate prices for CTs and CMs. I understand that wireless carriers do this because (among other reasons) if they were to purchase CTs and CMs from separate vendors, they would incur greater difficulties in fault-tracing/restoration when there is an outage on a circuit provided by more than one carrier.(41)
  8. Sprint Nextel uses CT and CM in combination to provide a circuit from a cell site to a wire center and then on to a more distant wire center or BOC central office. As a result of these complementary purchases, even when the BOCs have Phase II pricing flexibility only for CMs but not for CTs, and even though Sprint Nextel purchases those services separately at separate prices, the BOCs may be able to set the unconstrained rate of CMs to extract the monopoly profit from CTs. Thus, the sum of the two prices that Sprint Nextel pays for CT and CM will equal the sum of the unregulated monopoly prices for both, and in that way the BOC can reap the monopoly profits from the sale of both CMs and CTs.(42)
  9. In summary, Sprint Nextel's experience and that of other special access customers indicates that the emergence of competition anticipated by the Commission under the current price flexibility triggers did not occur, and today competition is not sufficiently robust to regulate the special access rates charged by the BOCs.

  1. Integrated BOCs have Incentives to Raise Wireless Rivals' Costs of Special Access
  1. CMRS carriers use special access purchased from the BOCs to link their cell sites to their own networks. At the same time these CMRS carriers also compete with the wireless affiliates of BOCs. A BOC that supplies CT and CM services to a CMRS competitor can raise that rival's costs by increasing special access rates in price flexibility areas where there are limited alternative sources of special access. By doing so the BOC's wireless affiliate benefits from weaker competition and an increased demand for its own wireless services.
  2. In the case of Verizon, partial ownership of its affiliate Verizon Wireless enables the BOC to realize a portion of the benefit of raising wireless rivals' costs through higher special access rates. In 2005, AT&T, which held a 60% interest in Cingular, was in a similar position.
  3. Now, as a result of its merger with Bell South, AT&T has full ownership of the Cingular (now AT&T Mobility) wireless operation. For a BOC that fully owns its wireless affiliate an increased price of special access is an internal transfer that does not affect the profit-maximizing calculation of its wireless division. As a result, AT&T's incentive to raise the costs of its wireless competitors, including Sprint Nextel, through higher special access rates has increased since 2005.

  1. The Exercise of Market Power: the Pricing Evidence
  1. The previous sections have illustrated the general absence of competition faced by the BOCs in the provision of special access. Thus, the structural conditions necessary for supracompetitive pricing exist in the markets for special access services. And it is the case that in areas where the BOCs have been granted pricing flexibility, the prices of special access services appear to have remained significantly above competitive levels. As I discuss below, that pricing behavior is evident from data that have been previously provided to the Commission.

  1. Comparing Rates Under Pricing Flexibility and Those Under Price Caps
  1. A meaningful analysis of rates requires comparisons that eliminate differences due to changes in spending on special access services that result from differences in volume purchased, duration of contract, and aggregation and bundling of services subject to different price regulations. Evaluating the effect of pricing flexibility on special access rates thus requires looking at a rate for a particular service for a particular term, with otherwise identical contractual provisions. The Commission has available to it the relevant comparisons from several sources.
  2. For example, in his Declaration in 2005 Dr. Joseph Stith, a mathematical statistician at pre-merger AT&T, calculated pro forma price flexibility rates (i.e., rates no longer subject to price caps), price-capped rates, and the UNE rate for each BOC in effect on July 1, 2004.(43) The pro forma rates provide a meaningful comparison by examining for the same commitment period exactly the same service offered for each capacity — a stand-alone circuit consisting of two channel terminations, a fixed mileage transport charge, a variable mileage transport charge, and an assumed 10 miles of transport.(44)
  3. While Dr. Stith's review shows that there is considerable variation in prices across BOCs, in no instance does the pro forma price flexibility rate for either a DS1 or DS3 comparison (for both month-to-month and 3-year optional pricing plans) fall below the pro forma price-capped rate.(45) When rates included in his review are averaged across all BOCs and all geographic areas, for month-to-month DS1 service the pricing flexibility rates are 19% above the price cap, and are nearly three times the UNE rates. For DS3s, the average month-to-month pricing flexibility rate is 15% greater than the price-capped rate and more than three times the UNE rate.(46)
  4. In her 2005 Declaration on behalf of Global Crossing, Janet Fischer charted service-specific comparisons between the BOCs' price flexibility and price-capped rates over the 1997-2005 period across various states and BOCs. The comparison for BellSouth reveals that as soon as pricing flexibility became available in 2001, the BOC immediately raised DS1 rates (for both CTs and CMs). And in three of Ms. Fisher's four comparisons for BellSouth, the price flexibility rate remained above the price-capped rate.(47) For the other BOCs examined by Ms. Fischer, the pricing flexibility rate remained constant or increased while the gap between the pricing flexibility rates and the price-capped rates increased.
  5. These results were updated and confirmed in 2006 by the Government Accountability Office (GAO). The GAO analyzed data on dedicated access from 16 MSA — four metropolitan areas in the geographic areas broadly served by each of the four BOCs (AT&T, BellSouth, Qwest, and Verizon).(48) It analyzed 1,152 list prices for CTs and CMs for monthly, 3-year, and 5-year term commitments and found that "as of June, 2006 the price-flex list price was on average higher than the price-cap price, regardless of whether the price was for channel terminations, interoffice mileage, DS-1 or DS-3 service, different term arrangements, or different density zones."(49) Adjusting prices for inflation using the producer price index for wired telecommunications carriers "does not change the result that prices are higher in Phase II areas on average, or that prices have increased over time in Phase II areas."(50)
  6. In order to capture the effect of contract discounts, GAO also compared average revenue for special access services in 56 MSAs — 27 MSAs with Phase II flexibility for CTs and 29 with Phase I.(51) As of 2005, average revenue in the Phase II areas is about 4 percent higher for DS1 CTs, and 24 percent higher for DS3 CTs, compared with average revenue in the Phase I areas.(52) Adjusting the average revenue data for inflation and weighting the data on the basis of the number of businesses in an MSA did not change the finding that Phase II average revenue was higher than Phase I average revenue.(53)
  7. Sprint Nextel has systematically compared special access tariffs in effect in July 2007 in Verizon territories. For a representative circuit consisting of two CTs and 10 miles of CM, Sprint Nextel's analysis confirms the patterns found in earlier studies. For DS1 circuits in the Verizon states, the month-to-month price flexibility rates are 21% to 33% greater than the month-to-month price cap rates, and the 5-year term price flexibility rates are 20% to 33% greater than the 5-year term price cap rates.(54)
  8. For DS3 circuits, Sprint Nextel finds that in the Verizon states the month-to-month price flexibility rates are 13% to 52% greater than the month-to-month price cap rates, and the 5-year term price flex rates are 6% to 42% greater than the 5-year term price cap rates.
  9. Pursuant to a commitment AT&T made to the Commission in seeking approval of its merger with BellSouth, in 2007 AT&T reduced special access rates in Phase II areas to the level of its rates in price cap areas.(55) Prior to this, AT&T's rates for DS1 10-mile circuits in 2006 were as much as 32% higher than the corresponding price-cap rates, and its rates for DS3 were also as much as 32% higher than the corresponding price-cap rates.(56)
  10. Sprint Nextel has also compared 5-year term price-cap rates for DS1 and DS3 10-mile circuits with UNE rates in AT&T and Verizon territories. Across five states (WI, TX, OH, MI, CA) AT&T's price-cap rates for DS1 circuits are from 53% to 248% greater than AT&T's comparable UNE rates. In four states (PA, NY, MA, MD) Verizon's price-cap rates for DS1 circuits are 24% to 126% greater than its comparable UNE rates. For DS3 circuits (except for a single density zone in California) the price cap rates for AT&T's DS3 circuits in the five states are greater than UNE rates, exceeding them by up to 165%. In the four Verizon states the price cap rates for DS3 circuits are from 4% to 59% greater than the UNE rates.(57)
  11. In summary, these data indicate that pricing flexibility has allowed the BOCs to increase prices above the levels that price caps otherwise would have allowed and far above the comparable UNE rates. Moreover, it appears that the onset of pricing flexibility was more likely to cause prices to increase rather than fall. This is hardly what one would have expected if the supply of special access services were effectively competitive.

  1. Contract Pricing Plans Enable BOCs to Extend Market Power
  1. Dr. Joseph Farrell, Professor of Economics at the University of California, Berkeley and a former Chief Economist at the FCC, showed that ILEC special access tariffs include pricing schemes that provide loyalty discounts, volume commitments and termination penalties that blunt the incentive for a customer to contract with a competitive carrier whose long-run cost is below the ILEC's price.(58) Dr. Farrell analyzed the incentives created by SBC's "Managed Value Plan" (MVP) Tariff. The MVP establishes a "Minimum Annual Revenue Commitment" (MARC) that the customer must maintain with the BOC for a five-year term in order to receive annual discounts that increase from 9% in the first year to 14% in the fifth year. He finds that the MARC combined with a termination penalty "sets up an automatic and sometimes drastic price cut for any portion of the customer's business that the customer is considering switching to a competitor."(59) Such pricing plans can have the effect of requiring a competitive carrier to beat a marginal price that is well below the average price that special access customers pay the BOC. As a result, special pricing terms can have the effect of excluding competitors unless they are able to enter the market on a large scale and serve all of the business of a sufficient number of customers.
  2. The GAO likewise found that many contracts that provide for discounts off the list price include revenue guarantees, requirements for shifting business away from competitors, and severe termination penalties. "Unless a competitor can meet the customer's entire demand, the customer has an incentive to stay with the incumbent and to purchase additional circuits from the incumbent, rather than switch to a competitor or purchase a portion of their demand from a competitor — even if the competitor is less expensive."(60)

  1. Summary
  1. The rigorous pro forma comparisons of special access prices that have been provided to the Commission (as well as those of the GAO) are consistent with a conclusion that the effect of pricing flexibility has been to allow the BOCs to exercise market power. Those comparisons reveal that the DS1 and DS3 rates charged by the BOCs exceed the equivalent price-capped rates and are far higher than one reasonable measure of costs of service, the comparable UNE rates. Moreover, the terms of special access contracts that the BOCs offer in price flexibility areas have the effect of excluding all but the largest competitors.

  1. Conclusion
  1. Previous sections have described some key "stubborn facts" that should lead the Commission to implement a more refined special access regulatory regime. The entry barriers to the provision of CT and CM services are substantial and only likely to be overcome in some parts of an MSA, not all of the MSA. The existence of those barriers has prevented special access competition from emerging throughout the MSA. And the effect of allowing the BOCs flexibility in pricing special access services in areas where competition is non-existent or nearly so has elevated special access rates further above competitive levels. As a result, consumers have been harmed through these higher prices.

I declare under penalty of perjury under the laws of the United States of America that the foregoing is true and correct.

Executed on August 8, 2007

Bridger M. Mitchell


1. Declaration of Bridger M. Mitchell and John R. Woodbury, filed as Attachment 1 to Reply Comments of Nextel Communications, Inc., WC Docket No. 05-25, (July29, 2005) (2005 Mitchell-Woodbury Decl. and 2005 Nextel Reply).

2. Special Access Rates for Price Cap Local Exchange Carriers; AT&T Corp. Petition for Rulemaking to Reform Regulation of Incumbent Local Exchange Carrier Rates for Interstate Special Access Services, WC Docket No. 05-25, RM-10593, Order and Notice of Proposed Rulemaking, 20 FCC Rcd 1994, ¶5 (2005) (footnote omitted) (2005 Special Access NPRM).

3. I use CT to refer to the circuit connecting the cell site to the serving wire center, and CM to refer to the circuit connecting the wire center to either another central office or to a point of interconnection. In a BOC tariff, the recurring monthly charge for a CM circuit is comprised of two Universal Service Order Code (USOC) rate elements: a fixed charge, and a per-mile charge for transport times the mileage between the wire center and the other central office or point of interconnection.

4. Direct comparisons of the prices of specific services for the different commitment terms, rather than the average revenue measures used by the BOCs, are the appropriate basis for drawing inferences about the extent to which the current pricing flexibility triggers have resulted in competitive special access prices.

5. Unbundled Access to Network Elements, Order on Remand, 20 FCC Rcd 2533, ¶¶78-80 (2005) (Triennial Review Remand Order or TRRO).

6. U.S. Department of Justice and Federal Trade Commission, Horizontal Merger Guidelines §1.0 (1992, as revised in 1997), available at: (Horizontal Merger Guidelines). Note that the implementation of this test assumes that special access prices are not already set at a monopoly profit-maximizing level. If prices were at the monopoly level, then any further price increase would be unprofitable, and so performing the test would then tell little about market definition, only that prices higher than those that maximize profits are less profitable.

7. The capacity of one DS3 is equivalent to 28 DS1s. TRRO ¶170. Because of this functional equivalence, the application of the hypothetical monopoly test to either CTs or CMs could lead one to a conclusion that for any particular geographic market, DS1 prices would ultimately constrain the prices of DS3s and vice-versa. Horizontal Merger Guidelines §1.0. Thus, one could conclude that there is a "high-capacity" CT market comprising both DS1s and DS3 services and similarly a "high capacity" CM market. Yet, for reasons explained in the text, that does not mean that the Commission should focus on only on the availability of "high capacity" special access services.

8. TRRO ¶¶ 166, 170-171.

9. TRRO ¶¶ 154, 170.

10. TRRO ¶181 n.490.

11. TRRO ¶170 n.469.

12. TRRO ¶171.

13. Id.

14. A key reason why stand-alone DS1s are not profitable is because at the level of traffic that they carry, the CLEC will invariably have higher unit costs than the BOC. The BOC already has numerous DS1 loops that are provided over copper or hybrid copper-to-fiber facilities. And where the BOC does provide DS1 loops over fiber, it can aggregate that DS1 traffic with other traffic to support deployment of a DS3 facility to the wire center. Thus, in the TRRO, the Commission described its transport impairment test as one that "examines the feasibility of duplicating dedicated transport facilities connecting incumbent LEC wire centers." TRRO ¶ 91.

15. TRRO ¶126 (citation omitted).

16. To be sure, the prospects for CM competition may be better than for CT competition. In contrast to the generally limited volume of traffic over a channel termination circuit, the volume of traffic that is aggregated and transported between two wire centers (or central offices) generally requires the capacity of one or more DS3s.

17. TRRO ¶¶ 79-80.

18. TRRO ¶152.

19. TRRO ¶155.

20. TRRO ¶¶ 150, 152-153.

21. Reply Declaration of Janusz A. Ordover and Robert D. Willig on Behalf of AT&T Corp., RM-10593, ¶29 (Jan.23, 2003) (Ordover/Willig Reply Declaration).

22. Id. ¶26.

23. Id. ¶29.

24. Id. ¶¶ 26-27.

25. The FCC's record in the TRRO established that "the cost of construction does not vary significantly by loop capacity" and "[t]he most significant portion of the costs incurred in building a fiber loop results from deploying the physical fiber infrastructure into underground conduit to a particular location, rather than from lighting the fiber-optic cable. ... [F]or these reasons, LECs do not typically construct fiber loop facilities at lower capacity levels, such as DS1 or DS3, but rather install high-capacity fiber-optic cables and then use electronics to light the fiber at specific capacity levels, often 'channelizing' these higher-capacity offerings into multiple lower-capacity streams." TRRO ¶150.

26. See Verizon Communications Inc. S-4 filed with the Securities and Exchange Commission (SEC) on April 12, 2005 at 44-5, available at: .

27. See AT&T Inc. S-4. filed with the SEC on March 11, 2005 at 26, available at: .

28. See AT&T Inc. S-4. filed with the SEC on March 31, 2006 at 26, available at: .

29. See, e.g., "Special Access 101," a presentation by US Telecom to Congressional staff on July 27, 2007, at page 5.

30. Declaration of Gary B. Lindsey, Attachment 1 to Comments of Sprint Nextel Corporation, WC Docket No. 05-25, ¶9 (August8, 2007) (Lindsey Declaration and 2007 Sprint Nextel Comments).

31. Id.

32. Declaration of Steven Sachs, filed as Attachment 2 to 2005 Nextel Reply, ¶¶10-11 (July29, 2005) (Sachs Declaration).

33. Declaration of Chris Sykes, filed as Attachment C to Comments of T-Mobile USA, Inc., WC Docket No. 05-25, ¶5 (June13, 2005) (Sykes Declaration).

34. Ordover/Willig Reply Declaration ¶30.

35. Reply Declaration of Dr. Lee Selwyn, ¶20, attached as Exhibit 3 to Reply Comments of AT&T Corp., RM-10593, Table 7 (Jan. 23, 2003) (Selwyn Declaration).

36. Id.

37. Id. Table 8.

38. Id. ¶ 48, citing Proceeding on Motion of the Commission to Investigate Methods to Improve and Maintain High Quality Special Services Performance by Verizon New York, Inc., Opinion and Order Modifying Special Services Guidelines for Verizon New York, Inc., Conforming Tariff, and Requiring Additional Performance Reporting, NY PSC Case 00-C-2051, at 7-8 (June 15, 2001).

39. Economics and Technology Incorporated, Competition in Access Markets: Reality or Illusion, Table 2.1 (August 2004) (ETI Competition in Access Markets), filed as Attachment A to Comments of the Ad Hoc Telecommunications Users Committee, WC Docket No. 05-25 (June13, 2005) (2005 Ad Hoc Comments). Updated data for New York and Washington metropolitan areas, Denver, San Francisco, Dallas, Oakland and Florida are consistent with this picture. Declaration of Susan M. Gately, filed as Attachment B of the 2005 Ad Hoc Comments, ¶¶ 17-19 (June13, 2005) (Gately Declaration).

40. Lindsey Declaration ¶¶4-5.

41. Sachs Declaration ¶ 7.

42. The BOC's ability to price in this way will depend in part on the extent to which special access customers tend to purchase CTs and CMs for joint usage, and the efficacy of any non-discrimination provisions.

43. Declaration of M. Joseph Stith, WC Docket No. 04-313; filed as Attachment H to Comments of AT&T Corp. (Oct. 4, 2004), Attachment 1 at 4, "Comparison of costs (10-mile Standalone Circuit) Rates in Effect on July 1, 2004" (Stith Declaration).

44. Stith Declaration ¶5. If a BOC has zoned rates, the calculation used the Zone 1 rate.

45. Stith Declaration, Attachment 1, at 1-20.

46. In order to calculate the differences in prices, a simple average of the percentage differences was taken in which each observation was the percentage difference between a state's rates as reported by Stith. In cases were UNE rates were not reported in Stith's tables, the observation was treated as a missing value and ignored in the UNE rate calculation.

47. Declaration of Janet S. Fischer on behalf of Global Crossing North America, Inc., attached to Comments of CompTel/ALTS, et al., WC Docket No. 05-25, Table 4 (June13, 2005). In the one exception, the price flexibility rate for DS1 month-to-month channel terminations converges to the price-capped rate because of an increase in the price-capped rate, not because of a fall in the price flexibility rate.

48. United States Government Accountability Office (GAO), Report to the Chairman, Committee on Government Reform, House of Representatives, Telecommunications: FCC Needs to Improve its Ability to Monitor and Determine the Extent of Competition in Dedicated Access Services, at 10 (Nov.30, 2006), available at: (GAO Report).

49. Id. at 28.

50. Id.

51. Id. at 62.

52. Id. at 32.

53. Id. at 62.

54. "Comparison of Price Cap and Pricing Flexibility Rates," Exhibit 1 to 2007 Sprint Nextel Comments (Exhibit 1).

55. See AT&T Inc. and BellSouth Corporation, Application for Transfer of Control, Memorandum Opinion and Order, 22 FCC Rcd 5662, Appendix F, Special Access Condition 6 (2007); Order on Reconsideration, 22 FCC Rcd 6285, ¶1 (2007).

56. Exhibit 1.

57. "Comparison of AT&T and Verizon Special Access Rates and Unbundled Network Element Rates," attached as Exhibit 3 to 2007 Sprint Nextel Comments.

58. Reply Declaration of Joseph Farrell, attached to Reply Comments of CompTel, et al., WC Docket No. 05-25, ¶ 4 (July 29, 2005).

59. Id. ¶ 11.

60. GAO Report at 30.

Updated June 25, 2015