Economics and Technology, Inc. -- Views and News, September 2011

September 2011

Continue reading ETI Views and News at econtech.com, or download a printer-friendly version.

Helping Themselves to Corporate Welfare: The Large ILECs' Plan for Universal Broadband

More than four decades have elapsed since the FCC embarked upon a mission to introduce competition into the US telecom market. Varying levels of competitive activity have emerged in most telecom sectors, with the notable exception of many rural, low-density areas that are thought to be incapable of supporting multiple providers. The FCC's March 2010 National Broadband Plan noted that at least two competing providers of high-speed Internet access were available to some 82% of the US population, and that for the 5% that still had no access to even one broadband network, high-cost support or other subsidies would be required in order to assure truly universal broadband availability nationwide.

Subsidization of telecom services in high-cost and insular areas has long been a major focus of US telecommunications policy. Initially this was accomplished by charging above-cost prices for certain optional and discretionary services (like long distance, access charges, and calling features) the "profit" from which would be used to support below-cost pricing of basic residential dial tone access and to provide additional subsidy to high-cost areas. The 1996 Telecommunications Act required these "implicit" subsidies to be replaced by explicit support mechanisms. These "Universal Service" funding programs were nominally focused upon voice services; the FCC had yet to extend explicit universal service support to broadband.

In February, the FCC initiated a rulemaking to address Universal Service and Intercarrier Compensation that was intended to consider, among other things, extending explicit universal service support to broadband. On July 29, six of the nation's largest ILEC/wireless/ broadband providers – AT&T, Verizon, CenturyLink, FairPoint, Frontier, and Windstream – offered what they described as a comprehensive plan for reform of universal service and intercarrier compensation, which they called their "America's Broadband Connectivity Plan" ("ABC Plan") for providing support for broadband infrastructure development in high-cost areas explicitly through broad-based contributions to several new funds to be established for this purpose. These seemingly ambitious goals aside, the ILECs’ "ABC Plan" is in reality a gambit aimed at protecting its sponsors’ subsidized monopoly status and embedded investments in legacy technologies that are actually at odds with the FCC’s stated goals.

The not-so-simple, not-so-equitable and not-so-efficient elements of the large ILECs’ so-called "ABC Plan"

The ILEC sponsors of the "ABC Plan" call it "a framework that ... will ensure that four million rural homes and businesses in high-cost areas served by price cap carriers will have access to broadband, two million of which will enjoy the benefits of broadband for the first time." Upon closer examination, however, it becomes apparent that the ILECs' plan stakes out a privileged position for its sponsors at the expense of smaller competitors and consumers. The Plan contemplates three distinct funds, totaling no more than $4.5-billion annually through at least 2022:

• the "Connect America Fund" ("CAF") for the large price cap incumbent LECs, funded at $2.2 billion;

• a separate fund for small rate-of-return rural LECs (initially funded at $2-billion but growing to $2.3-billion); and,

• (if funds are available) an "Advanced Mobility/Satellite Fund" ("AMF") capped at $300-million annually for satellite or other wireless broadband in what would otherwise be extraordinarily high-cost-to-serve-by-means-of-wireline areas.

CAF funding would be available for broadband service that provides a minimum actual downstream and upstream bandwidths of 4 mbps and 768 kbps, respectively, and that can be furnished using "any wireline or wireless technology." CAF support "is only available in those high-cost areas in which there is no private sector business case to offer broadband." However, the Plan contains no actual mechanism to evaluate any such "business case," relying instead upon a presumption that wherever there was no "unsupported broadband competitor ... already offering broadband service as of January 1, 2012 in the "high cost" areas, then that must mean that there is no "business case" to be made for such entry. In other words, what the Plan deems to be a "business case" is essentially the actual presence of a competing non-ILEC provider. The ABC Plan freezes the determination of support as of a date certain (January 1, 2012); "the entry of an unsupported broadband competitor after January 1, 2012 does not affect the level of CAF support."

Once an area is determined not to have at least one unsupported competitor, the next step is to determine whether it qualifies as "high-cost." For this purpose, the ABC Plan sponsors commissioned CostQuest Associates, Inc. to develop a cost model that calculates the forward-looking cost of providing broadband (and the estimated support levels) separately for each individual census block based upon the use of wireline technology. Under the ABC Plan, a high-cost census block is one where the average per-served-location monthly cost exceeds a threshold of $80. While costs are modeled at the census block level, support is determined and disbursed by aggregating all of the eligible (high cost and unserved by an unsupported competitor) census blocks within a wire center, referred to as a "supported area" of the subject wire center.

Census Blocks and Wire Centers

The basic analytical element of the ILEC Cost Model is a geographic area known as a "census block." A "census block" is, as its name suggests, a construct of the Bureau of the Census whose purpose is to serve as a basis for data collection and organization, certainly not for designing telecommunications infrastructure. "Census blocks" represent the smallest geographic area for which the Bureau of the Census collects and tabulates decennial census data. They are formed by streets, roads, railroads, streams and other bodies of water, other visible physical and cultural features, and certain legal boundaries such as municipal or county lines. While this "census block" approach for modeling telecom costs has been promoted for its "granular" focus, the purported benefit of using these small, but essentially arbitrary and, from a network engineering standpoint, utterly meaningless, geographic units ignores the fact that census blocks have no relationship to broadband deployment costs and the resulting investment decisions.

Networks are not designed around – and network costs are not incurred at – the census block level. While the precise manner in which the ILEC Cost Model assigns costs to individual census blocks is not apparent from the available documentation, if the model is actually calculating the stand-alone costs of serving individual census blocks, it will necessarily exaggerate these costs by ignoring scale and scope economies extant across larger geographic areas. And if it is calculating costs across geographic areas larger than individual census blocks, there is no assurance that costs common to multiple blocks are being properly assigned and attributed to the individual census blocks that would be the unit for receipt of high-cost support.

Not only are the criteria used to define census blocks basically irrelevant to how networks are designed and cables routed, these criteria can often be directly incompatible. For example, census blocks are typically bounded by streets, roads or other public ways, such that customers on either side of any given street or road will normally fall within different census blocks. But both sides of a street will typically be served by the same wireline distribution facility. One can even imagine a situation where, under the ILEC Cost Model, the census block on the west side of the road falls below the threshold for CAF support, while the modeled per-location cost for the block on the east side of the same road is above the threshold and thus qualifies for CAF support, even though customers on both sides of the road are served from the same common distribution cable. And if, as a result of the competitive bidding process that the Plan calls for in certain circumstances, the provider responsible for the "high cost" side of the road ends up not being the same as the one that serves the "low cost" side, the economic benefits of serving both sides of the road from the same distribution facility could be sacrificed. Moreover, were that to happen, the cost of serving the "low cost" side of the road could well escalate, perhaps even placing that census block above the support threshold.

Compounding the ILEC Cost Model’s misuse of the census block as a basis for modeling a telecommunications network, the Model is integrally linked to legacy wireline – and specifically, ILEC – network architecture, i.e., the wire center. It employs what is frequently referred to as "scorched node" network design in which legacy ILEC wire center buildings and serving areas are maintained as they presently exist. Many, perhaps most, of these wire center locations and serving areas date back to a time when transport costs were high and large digital switches did not exist. Anyone designing a broadband network from scratch – i.e., a so-called "greenfield" build – would adopt a "scorched earth" approach with no preexisting location, configuration, technology or network architecture constraints. Past limitations on transport distances, telephone switch and inter-switch trunk capacities, and other attributes of legacy voice telephony technology were materially responsible for dictating the design of local networks. Fiber optics, packet switching, wireless and other current technologies – together with the evolving demand for increased bandwidth, mobility, and applications that go way beyond point-to-point voice telephone calls – fundamentally change the way a network would be designed from the ground up today and in the future.

CLEC, cable, and wireless networks are not oriented around the limited geography embraced by legacy ILEC wire centers. Were a forward-looking cost model based upon geographies as small as wire centers – or worse, the even smaller individual census blocks – used as a basis for modeling CLEC or wireless costs, they might well appear to exceed the costs associated with traditional ILEC networks because they would ignore the significant efficiencies associated with modern wireline and wireless network architectures that are oriented around far more expansive geographic service areas. Thus, even if the ILEC Cost Model did not deliberately exclude wireless solutions, if wireless were costed on a census block basis – something that would literally never happen in the real world – the per-location costs for a wireless solution based upon the legacy ILEC wireline network architectures assumed by the ILEC Cost Model could well appear to be higher than for wireline.

The use of wireline-only costs results in skewed and potentially wasteful support decisions

The ABC Plan claims that "[t]he broadband service obligation is technology-neutral: providers can use any wireline or wireless technology that meets the specified bandwidth and service requirements." Upon closer examination, however, this putative "technology-neutral" aspect of the proposal is undermined by several decidedly ILEC-oriented aspects of the Plan and the ILEC-commissioned cost model that underlies it. For one, the cost model considers only wireline technology as the strategy for providing broadband at the level of geographic intensity contemplated in the National Broadband Plan. In adopting this constraint, the Model makes no attempt to identify or assess the most efficient, least cost means of providing broadband service in unserved (typically low-density, rural) areas being specifically targeted by the FCC’s Broadband initiative – the very locations where wireless offers significant cost advantages over wireline. With the exception of the most extreme high-cost areas (again, as determined under the assumption of wireline deployment), the ILEC Cost Model simply assumes that a wireline solution is superior to wireless both with respect to cost and, apparently, in its ability to achieve the requisite 4 Mbps download/768 Kbps upload minimum service objective. In fact, wireless has unique cost advantages in low-density geographic areas that are overlooked entirely by the wireline-oriented ILEC Cost Model:

• Per-location costs of serving customers in low-density areas via wireline distribution networks are high due to the confluence of large distances and the small number of locations to be served. The costs of wireline distribution facilities – "last mile" subscriber lines and supporting structures (poles and conduits) are primarily driven by distance and by terrain and, to a much lesser extent, by total capacity of the distribution facility. Holding capacity constant, a six-mile distribution facility costs roughly twice as a three-mile facility; holding distance constant, a distribution facility capable of serving 1000 locations costs little more than a distribution facility capable of serving 200 locations, the principal source of difference being the cost of the coaxial or fiber cable itself. As distance increases while density becomes more sparse, the costs of serving customers via wireline broadband escalates rapidly.

• Wireline drops also tend to be most costly in many rural areas where the distance from the street or road to the subscriber's residence may often be considerably longer than in urban and suburban areas – and may be more costly to maintain on an ongoing basis. With wireless, there is no need to construct a drop cable from the road to the house, since the wireless service can be received directly at the customer's residence.

• Terrestrial wireless technology is often particularly well-suited to low-density rural areas. First, rural areas generally do not face the same level of spectrum congestion extant in areas of greater density. Second, in many low-density areas – particularly where the terrain is relatively flat – a single cell site can serve a considerably larger area than is typically possible in urbanized or even suburban areas. Thus, while the wireless cost per location served is still somewhat higher in rural areas than in urban/ suburban communities, the differential between the two extremes is likely far smaller than for wireline.

The wireline bias inherent in the ILEC Cost Model would produce technologically inefficient results even if it were merely proposed as the basis for identifying census blocks with (wireline) costs above the designated ($80 per month) threshold. But the ABC Plan also proposes that the model's results will establish the specific level of "baseline support," by individual census block, throughout the sponsoring ILECs’ service territories. The overstatement of "forward looking" costs that would result from a cost model that expressly excludes consideration of a potentially lower-cost technology, coupled with the presumptive incumbent LEC "right of first refusal" bias, will bloat the aggregate level of support to be provided under the CAF mechanism and, since the aggregate level of CAF support will be a major determinant of the aggregate level of explicit contribution to be made to the Fund, will result in excessive prices for all services that are to be subject to such contribution requirements, which will in turn have broad negative impacts on the economy overall.

A Fundamentally Wrong-headed Policy Framework

Adoption of the "ABC Plan" would be a serious step in the wrong direction. It would from the very outset deny rural consumers access to a broad range of mobile broadband services. It would lock in for a decade or longer a fixed-location broadband service standard that is barely adequate to support current needs, and that will almost surely fail to keep up with new demands over the lock-in period. It will result in an excessive level of support by limiting its cost modeling to legacy wireline technology and network architecture, thus ignoring the least-cost forward-looking solution. And, by preempting wireless (and other) providers' ability to compete for support in nearly five-sixths of all supported locations, the ABC Plan will place an excessive economic burden upon all services and geographic areas that are being required to contribute such support. In the end, the large ILECs' ABC Plan does not resolve the universal service/intercarrier compensation reform question; rather, it largely kicks it down the road for another decade – and still provides no assurance that even then we will achieve the needed outcome. Here’s what the FCC needs to do:

• Require that the maximum support available to any area be capped at a level consistent with the least-cost forward looking technology.

• Establish "supported area" geographies that offer efficient service opportunities based upon current and forward-looking network technologies and design standards.

• Include broad availability of mobile as well as fixed-location broadband services in the objective minimum service standard, rather than (as the ABC Plan would do) limiting support for mobile services to areas with extraordinarily high cost.

• Eliminate all "right of first refusal" or other preemptive claims on support, making every "supported area" open and available to competitive bidding.

• Adopt a minimum service standard that has the flexibility to evolve with changing demands for broadband capabilities, and include as a condition for eligibility a potential support recipient's ability to make ongoing adjustments in its service mix over the support period.

• Eliminate "make whole" and "replacement" revenue arrangements that reward incumbent carrier inefficiencies by shifting cost burdens to areas and services facing the fewest competitive alternatives, that frustrate competitive entry, and that distort competitive choices where available.

Deploying broadband in rural and high-cost areas is a costly undertaking by any standard and, to the extent that those costs will be borne by the rest of the economy, it is critical that this be accomplished as efficiently as possible. The ABC Plan expressly excludes potentially lower-cost solutions and forecloses participation by potentially lower cost providers, thus bloating the overall funding requirement. Excessive spending in this pursuit creates a drag on the economy generally and needs to be avoided. The ILECs' Plan is simply not ready for serious consideration, and any attempt by its sponsors to rush it through the FCC's decision process should not be condoned.

For more information, contact Lee L. Selwyn at lselwyn@econtech.com

Read the rest of Views and News, September 2011.

A more detailed examination of the large ILECs' "ABC Plan" as submitted in FCC WC Docket No. 10-90 is provided in ETI’s new report, "The Price Cap LECs' ‘Broadband Connectivity Plan’ – Protecting Their Past, Hijacking the Nation's Future".

Available at www.econtech.com/pubs.php

About ETI. Founded in 1972, Economics and Technology, Inc. is a leading research and consulting firm specializing in telecommunications regulation and policy, litigation support, taxation, service procurement, and negotiation. ETI serves a wide range of telecom industry stakeholders in the US and abroad, including telecommunications carriers, attorneys and their clients, consumer advocates, state and local governments, regulatory agencies, and large corporate, institutional and government purchasers of telecom services.