TVA Rural Studies

Improving Rural Telecommunications Infrastructure

Bruce L. Egan
Columbia University

7. Infrastructure Development

Now and in the foreseeable future, federal state and local governments play a key role in developing the rural telecommunications infrastructure. Indeed, regulators are largely responsible (along with the industry) for creating the current complex web of industry cross subsidies which are the very lifeblood of many rural systems and which allow rural POTS subscribers to enjoy a level of service and prices that are at par with urban and suburban subscribers.

As the industry transitions from a monopoly structure to a competitive one, rural subsidies are clearly at risk. The federal government remains concerned about rural issues and both major pieces of federal legislation aimed at furthering competition in the industry contain provisions to protect subsidies for universal service including both low income subscribers and rural high cost areas.35

7.1 RUS Guidelines for Borrower Companies

The RUS provides low interest loans for network upgrades to the majority of small rural LECs. As the primary source of public funding for rural telephone network upgrades, the RUS’s published guidelines for network upgrades has important national standing for infrastructure policy. The most recent guidelines were adopted on March 15, l995. Its major provisions are:36

  • every State must have a modernization plan to improve the rural telecommunications network and must submit it for RUS approval (may be drafted by the State regulatory agency or the borrower companies themselves);
  • the plan must provide for the elimination of party line service;
  • the plan must provide for availability of services for improved business, educational, and medical services;
  • the plan must encourage and improve computer and information highways for subscribers in rural areas;
  • the plan must provide for rural subscribers to receive
    • conference calling
    • video
    • data rates of at least 1 Mb/s;
  • uniform deployment schedules in rural and nonrural areas;
  • expeditious deployment and integration of emerging technologies;
  • affordable tariff rates for medical and educational services;
  • reliable powering for POTS service including alternative power sources during electric utility power outages;
  • in the "short term" all new telecom network facilities shall be constructed so that all single party service subscribers have access to
    • lines capable of speed of at least 1 Mb/s
    • switching equipment that supports custom calling features
    • E9ll;
  • in the medium term (6 years after plan approval) all new facilities must be capable of
    • transmitting (motion) video signals
    • E911;
  • in the long term all plans should accomplish
    • an elimination of party line service
    • universal availability of digital voice and data service (56-164 kb/s)
    • service at transmission speeds of no less than 1 Mb/s
    • video service

Needless to say, the new RUS guidelines and the considerable network capabilities which are required have sparked equally considerable controversy.37 Suffice it to say that while the RUS has laid out the rules for approving loan applications for network upgrades, it is far from clear that there is enough money available to pay for such substantial upgrades and it equally unclear whether or not non-borrower companies would otherwise plan to make such upgrades. A consistent state infrastructure upgrade policy would ideally be based on an industry consensus, but with the RUS setting the least common denominator at such a high level it may not be possible to reach an industry consensus. Keeping in mind that many large LECs serve rural areas too (in fact most rural areas of the US are served by large LECs), until the large LECs (which cannot borrow from the RUS) concur in the RUS proposals (not a likely proposition) it will not be possible for States to implement consistent network infrastructure upgrades plans.

It appears that the RUS rules were written for an era of continued monopoly provisioning of telephone service. This model is outdated in light of other federal and state initiatives promoting market entry. Large and small LECs, whether or not they are RUS borrowers, recognize that the future competitive environment means that local network upgrades involve considerable market risk, and there is no longer any good prospect of recovering all of the investment from tariff rates on captive customers. One obvious provision is lacking in the RUS's rules - that infrastructure investment plans meet the fundamental test of market viability (i.e., there is no business case called for).

This having been said, the RUS should be applauded for its vision, which, rather than a mandate, is a reasonable goal to strive for. The problem is that without large LEC concurrence it will never be implemented on a large scale. The RUS probably sensed this when they called for state PUCs, which regulate all LECs both large and small, to coordinate and submit their own infrastructure upgrade plans in accordance with RUS minimum requirements.

It will be up to the state and federal governments to try to coordinate their respective roles regarding telecommunications infrastructure policy. This is not an easy task and it has barely begun.38

7.2 Public Power Grid

The use of a dielectric transmission medium, such as fiber optics, provides an unprecedented opportunity for inexpensive infrastructure development by taking advantage of the new-found synergies of combining existing electric utility distribution infrastructures with those of telecommunications. Construction costs of fiber optic facilities may be substantially reduced by utilizing public power grid rights of way and pole or conduit facilities. Since optical transmission is not susceptible to the electromagnetic interference caused by power lines, fiber cables could use the distribution plant offered by the statewide power grid by purchasing or leasing facilities from rural electric utilities. Such inexpensive fiber deployment may even include lashing the fiber cable to the electric utility ground and phase wires, which often run along the tops of towers and poles. There are many possibilities. One new product on the market is a fiber cable which utilizes the metallic ground wire for strength. The ground wire in the cable supports the requirements of electric utilities, and the fiber communications capacity may be resold.

Power companies are heavy users of communication services, and many large utilities already operate major private communication networks. Smaller rural electric companies also require communications for load management, monitoring, internal communication and the like. Rural Electric Cooperatives serve geographically large and thin rural markets which often span many independent telephone company exchanges. Because they cannot justify "stand alone" internal communication networks, small electric utilities must rely on many rural telephone companies, and pay relatively high tariff rates. The sharing of power company facilities with local telephone companies can provide economies for both, providing a "win-win" situation. In addition, some large businesses who choose to locate in rural areas are often able to get sufficient power, while advanced communications capability is lacking. If the shared infrastructure were available, businesses might be more likely to locate and expand in rural areas. Safety communications for fire and alarms are other new service applications which place only nominal bandwidth requirements on the communications infrastructure. There seems to be a natural synergy here for rural communication infrastructure development, but one that is under-exploited. The electric power industry tends to be very conservative, but many firms are now examining novel arrangements with communication service providers.

7.3 Toll Service

Rural telephone companies, long desiring direct entry into the lucrative toll market, could begin taking advantage of the revenue opportunities that a fiber optic infrastructure could provide, including the possibility of providing new data and video services. This is very important if the traditional large telephone company toll subsidies enjoyed by small rural telephone companies truly disappear due to increasing competition. Small rural telephone utilities may pool traffic and interconnect with the fiber optic backbone trunk network to efficiently, and profitably, provide high quality toll voice and data services. Fiber optic backbone networks may also allow rural subscribers to purchase digital services and access remote databases of enhanced service vendors.

7.4 State Planning

The process of rural telecommunication infrastructure development is an evolutionary one that will occur only gradually as advanced facilities become available. For this reason it is important that the process begin as soon as possible. State telecommunication planners must take on the role of coordinating network interconnection and development activities, exploiting potential synergies for the benefit of all subscribers In the early stages, such coordination will concentrate on surveying all of the communication facilities, public and private, and evaluating short and long term interconnection and compatibility potential. At first, microwave and satellite network facilities will be evaluated, along with existing coaxial cable network facilities, to determine interim infrastructure possibilities. The long term focus will be on migrating to a more efficient infrastructure based on digital fiber optics and radio technology; the goal will be to share network facilities whenever it is cost effective to do so, and guide the replacement of older network facilities with advanced facilities, stressing network compatibility along the way. Without compatibility, interconnection of communication networks will be inefficient or even impossible, and potential synergies are lost.

The rate of development for rural telecommunication infrastructures may depend largely on demand drivers. There are some logical ways to pursue network technology adoption, paying close attention to demand patterns in the current infrastructure. For example, secondary and tertiary schools, libraries, hospitals, and regional airports tend to be among the heaviest consumers of information and telecommunication services in rural areas. Public power utilities and other rural infrastructure firms, including occasional large manufacturing or service companies, also represent logical node points for rural networks. Existing telephone company switching offices, combined with the aforementioned, represent demand drivers and potential network hub sites, providing for efficient communication infrastructures. This set of candidates for network node (hub) points should generate a number of alternative deployment scenarios for state telecommunication planners to consider. Hubbing allows the economies of satellite, microwave, and fiber transmission to be used cost effectively in relatively thin markets, thereby maximizing the net present value of the rural construction program.

7.5 Regulatory Issues

Planning for an advanced rural telecommunication infrastructure raises many regulatory and public policy issues. Prominent among them is: Who should own and control the infrastructure and how should it be financed? There are obviously no "right" answers to such questions, but some general economic principles may guide the thinking on these issues First, private ownership and control is generally preferred to public ownership and control, for reasons of operating efficiency incentives that competition provides.39 Second, government must have a pro-active role as an overseer, enabler, and planner. As discussed previously, private network development may help support infrastructure development in a "win-win" situation where net revenue opportunities accrue to both private and public network participants through efficient interconnection and compatibility. The role of state government may be most helpful in identifying where public and private communications network activities may complement one another and strengthen the overall infrastructure.

As a rule, an infrastructure approach does not imply centralized ownership or control. It does imply cooperation among the various players, however, and this is the enabling role of government bringing together the players and encouraging infrastructure development. Much more can be done than we observe today. Most states have not yet placed sufficient emphasis on telecommunication infrastructure and its role in economic development, even in rural areas. New technologies just beginning to be deployed have very low unit costs once demand thresholds are met, but have very high up front capital costs. For this reason, an infrastructure approach to planning, which maximizes capacity sharing through a "hubbing" network architecture, holds great promise for dealing with the problem of thin rural markets. For example, even in Kentucky, which is considered a rural state, there are many locations which could generate enough traffic demand to justify a fiber, radio or satellite hub, depending upon the specific demand application(s) required.40 Eventually fiber hubbing would dominate as the technology of choice for most new shared network applications, while microwave radio, coaxial and copper cable will be used for dedicated short haul subscriber plant; with satellite and microwave radio utilized whenever wireline facilities cannot be deployed cost effectively, especially in physically remote applications.

Finally, there are a host of important pricing issues associated with recovering the costs of advanced telecommunications infrastructure development. Two primary issues, are broad toll rate averaging across the nation and toll-to-local service subsidies. Trends in both of these areas are troubling for rural telephone companies, and will no doubt become the subject of extensive public policy debates. A full discussion of these issues is beyond the scope of this paper but a few observations deserve brief discussion.

Increasing competition in toll services, and the absence of regulatory rules for retail tariffs of competitive toll carriers is slowly eroding the broad rate averaging rules which have been in effect for many years. The effect of rate averaging is to subsidize subscribers in thin rural markets relative to those in dense markets. New volume discounts for heavy toll users, especially business customers, have already been undermining traditional rate averaging. Regional rate de-averaging is likely to occur eventually. The toll subsidy which generally flows from larger telephone companies to smaller ones is also going to decrease as competition continues to drive prices down. The best solution here is probably to target subsidies more carefully toward only those companies who are most in need rather than toward entire classes of small companies as is currently the case.

7.6 Rural Telephone Service Subsidies

Using REA data as a representative proxy for high cost rural areas of the US, the total rural subsidy is estimated to be $5B per year based on revenues of $14B and costs of $19B.41 This means that rural rates on average would have to rise about 35% in order to pat the full costs of providing rural POTS service. This means an increase in average monthly tariff rates of about $19.

But this is a broad average figure is substantially understated if open competition were allowed. Competitive entrants will tend to pursue only the profitable rural LEC subscribers, like those which reside in relatively dense downtown areas of the rural exchange, larger business subscribers, and those subscribers who purchase a lot of non-basic services. These subscribers also provide substantial contributions toward funding the high costs, and supporting the low average tariff rates for the high cost, low (or no) profit subscribers. Without individual subscriber data on costs and corresponding revenues it is not possible to estimate what the cross subsidy flows within the rural LEC exchanges would be, but it is well known that the most profitable subscribers who are responsible for generating the most revenue, provide an inordinate amount of funds to support the costs of the remaining majority. Thus it is easy to see that the rural subsidy to truly high cost unprofitable subscribers who need to be much more heavily subsidized, is at least two times or more than the $19 per month quoted earlier.

If competition and subsidies for small LECs are to coexist, then there are some general economic principles which should underlie the subsidy funding mechanism. The mechanism should be:

  • fundamentally fair for consumers;
  • competitively neutral for competitors; and,
  • long term sustainable.

Furthermore, pursuant to a host of recent federal and state investigations of Universal Service it appears that almost all parties to the debate have agreed on certain aspects of various plans and proposals:

  • All charges for telephone service should be cost compensatory except for a narrowly defined and highly targeted set of basic service subscribers (e.g., low income, high cost).
  • Subsidy funding requirements should be shared by a broad base of market players and the funding mechanism should be administered by an independent third party.

Beyond the need to define exactly what constitutes the benefited service and subscriber group deserving of subsidies, there are some primary unresolved issues:

  1. What is the cost of the current and future public service obligations (e.g., Universal Service Obligation (USO) and Carrier of Last Resort (COLR)) and how should the cost be determined?
  2. Exactly how should the subsidy funding mechanism work? (e. g., How to administer the subsidy fund? Who should administer the fund? How to collect and distribute funds within any given geographic area—local, regional or statewide.)

Assuming, for purposes of discussion, that subsidies should only apply to covering a portion of the costs of providing Plain Old Telephone Service (POTS) for certain residential PSTN subscribers (e.g., high cost/low income), and assuming also that all other subscribers should directly pay the cost of their POTS service, then it becomes clear that the proper cost “object” for quantifying a subsidy requirement for USO and COLR is the total PSTN investment and expense cost of the LEC minus the cost of the (hypothetical) PSTN if there were no such obligation to serve or be ready to serve. (Put it another way, had the LEC not had the social compact with the certification authority that it would be the only service provider and that it would be allowed to recover its total costs from tariff rates charged to its subscribers what would its investment and expense cost be.)

A cost study method must be developed which establishes both the historical and going forward costs of the USO. In particular, the total (or average per line) cost of the USO is simply the cost of providing POTS service to residential subscribers. When this cost is compared to POTS revenues, the difference is the net cost (subsidy requirement) of the obligation to serve.

In the current local monopoly environment, the funding of this subsidy amount comes from numerous sources, namely, mark ups of prices over costs (i. e., cross subsidies) for LEC business services and residential non-POTS services. In the new competitive environment, all carriers must be required, in a competitively neutral way, to share in the funding of the costs of public service obligations.

Historical Perspective

The reason that the historical cost of the LEC's public service obligations must be considered is that the monthly subscriber bills now and in the future simply represent the time payment plan that regulators (on behalf of PSTN subscribers) and the LECs agreed to under the pre-competitive regulatory regime. Thus, in the post-competitive regulatory environment, the LEC must still be enabled to reimburse its historical investors. Had the LEC and its investors known that the regulatory regime would change such that it would not have the future opportunity to compete at the margin and still recover the cost of its regulatorily imposed historical public service obligations, then the LEC would not have kept investing in network infrastructure to certain customers and locations which it viewed as too risky without good prospects for cost recovery.

Incremental Perspective

The definition of the going forward cost of a LEC's public service obligations is the same as the historical one except that the costs would be based on the total incremental cost of the obligation instead of the total historical cost. Thus, the cost calculations would reflect incremental instead of embedded technology and business practices.

LRSIC vs. Total (Average) Cost

Some parties to the subsidy debate have asserted that the cost and funding of LEC public service obligations should be based on the incremental cost of POTS, not the total or average cost. Unless the Long Run Service Incremental Cost (LRSIC) study methodology they espouse is carefully constructed to approximate the total cost of the obligation; then it is flawed. This is unjustifiable from an economic perspective. This theoretical and practical distinction is important since it is the total (average) cost of a business enterprise which must be covered by total (average) revenues for the firm to be sustainable. Open market entry and competition will, over time, force rates for all services to be driven toward their costs (unless subsidies are provided).

Thus, a subsidy system which relies on continued price cost margins on competitive services to fund a portion of the costs for residential POTS service is an inherently unsustainable system. Competition is the natural enemy of cross subsidies. Indeed, a LEC, and, in turn, the LECs customers, which must incur the costs of regulatorily imposed public service obligations which are not similarly borne by market entrants (and their subscribers) will be disadvantaged in the market place The end result is a shift in consumer welfare from the LEC’s customer base toward the entrant's subscribers.

To avoid such discrimination and inadvertent shifts of wealth between subscriber groups, regulators should adopt a consumer friendly and competitively neutral subsidy funding mechanism before competitive entry forces a de facto rate deaveraging where one group of subscribers must pay for regulatorily imposed obligations and one does not. Fundamentally, there are only two types of POTS subscribers: those low cost subscribers in relatively dense areas served by relatively short loops who have relatively low local calling rates (subsidizers); and those relatively high cost subscribers served by relatively long loops which have relatively high local calling rates (subsidizees). Of course, even in high cost (long loop) geographic locations there will almost always be highly profitable subscribers who purchase a lot of non-POTS services. For this reason, a proper cost analysis must be conducted at the individual customer level. In other words, classifying a particular subscriber as a net subsidizer or a net subsidizee requires monitoring individual subscriber characteristics. Obviously, this is not practical even though it is the only way to guarantee that subsidies flow to subscribers who would otherwise not be able to obtain POTS service (or at least that quality of POTS which "profitable" PSTN subscribers receive).

Because LECs were historically monopoly providers of residential POTS, it is straightforward to calculate the total cost of any given LEC’s public service obligation by examining the embedded accounting cost data for PSTN investment and expense and converting it to an annual or monthly per subscriber amount.

In a historical context, LECs, as regulated common carrier monopoly providers of POTS, were obligated to serve (or be ready to serve) all subscribers on demand in a given service area at broadly averaged affordable POTS prices. Thus, the historical cost of the obligation to serve would be the cost a LEC would have incurred. The average subsidy requirement for rural telephone subscribers in the US was previously provided. Even this broad gauge level of analysis provides the regulators with a "ballpark" estimate of the total and average per subscriber subsidy. It is important to keep in mind, however, that this broad gauge estimate cannot be used to evaluate the cost (subsidy requirement) for any given subscriber or location because the available information is not sufficiently granular to make that determination. The average is the net total subsidy requirement from all POTS subscribers; the sum of those subscribers which are subsidizers and subsidizees. Nevertheless, the estimate does establish the total (average) amount of the subsidy funding requirement for the (as yet to be determined) subsidy recovery mechanism.

The total and average subsidy amounts determined in stage one of the costing process may begin to be deaveraged into density cells by examining the available data for a LEC's POTS subscriber density characteristics. This Phase is made somewhat easier because it is also possible to examine the PSTN investment and expense costs of smaller rural (less dense) LECs and comparing that to the same embedded accounting data for larger (more dense) LECs (see Table 4). The results from density cell analysis may be further disaggregated down to customer specific analysis via a computerized cost "proxy" model which calculates the costs for each subscriber location.42

7.7 Funding Mechanism

The mechanism for collecting funds to pay the cost (subsidy requirement) of LEC public service obligations must be competitively neutral with respect to incumbent LECs, which have the obligations, and competitive entrants, which do not. In other words the contributions toward covering the subsidy costs should be shared equitably by all telecommunication service providers. In order to be sustainable, no service providers should be able to avoid payments via bypass of the LECs PSTN. Many such funding mechanisms have been proposed by industry groups and most involve a revenue surcharge mechanism.43

7.8 Financing Alternatives

The costs of the deployment of efficient communication infrastructures are high compared to any historical measure of the costs of technology adoption. The reason is that the technological trends are toward lower on-going usage costs, and higher up front capital costs. Digital network equipment has few moving parts and features very large scale capacity relative to older generation network equipment. As such, the new equipment is more cost efficient from a maintenance and repair expense perspective, but is more capital intensive and is typically purchased in greater "lumps", because it is well suited for large scale operations. The same tends to be true of fiber optic transmission equipment, although for many network applications fiber will soon be cost effective even relative to the older generation copper and coaxial cable costs. The bonus with fiber optics is not only its very high capacity, but also its high quality and reliable service as compared to metallic and radio technologies. Nevertheless, up front deployment costs for fiber optics are substantial, and every effort to cost it effectively is important.

Telephone rates are the obvious first choice for financing advanced rural network infrastructures. Indeed, most of the financing must come from this source. Fortunately, it appears likely that the internal capital flows of telephone utilities will fund much of infrastructure deployment costs. But as was pointed out previously, these traditional internal cash flows are at risk due to increasing competition and the advance of technological alternatives.

Borrowing is the next alternative to consider. The U.S. Department of Agriculture's Rural Utilities Service and others provide subsidized loans to rural telephone companies. Without government assistance these telephone companies would have to go to other capital markets that offer less attractive terms.

Unlike large telephone utilities, many rural companies are already highly leveraged. This is not bad in and of itself, but it does impact the propensity of lenders to approve more funds on favorable terms. Regulators may also become concerned about the level of business risk which leverage implies, even though ratepayers may benefit from the lower average cost of debt capital relative to equity finance.

The RUS and some other lender's practices are basically sound for financing advanced rural telecommunication infrastructures because they operate within an incentive structure which tends to signal borrowers to make good investments (not to mention the new aggressive network upgrade guidelines). The RUS uses "equity based" financing and loans that are usually “self liquidating”. The proposed investments of borrowers must meet general technical guidelines for acceptable and approved equipment purchases. This system prevents speculation and abuse of government loan funds. Even though the RUS program is a loan subsidy program, only the interest rate discount is truly "subsidized”, and this is a relatively small portion of the entire loan and repayment sum. The loans are self liquidating from revenue and cash flow from telephone rates. Overall this approach seems socially efficient since it allows the private sector to determine the market requirements and opportunities for sound investment decisions, and requires the borrower to have a substantial equity stake. The only government role is to provide an inexpensive source of funds, technical support, and monitoring.

Direct subsidies, especially of the current untargeted variety, are much worse and are often not socially efficient. The current flow of toll-to-local subsidies from many large telephone companies to many smaller ones is generally inefficient because it is not based on need; instead, it is based simply on a grand formula for broad rate averaging and revenue sharing. In fact, some of the vast sums of money in the toll revenue pool now divided among telephone companies through the use of a broad formula could be used to increase the RUS’s loan authority or could be distributed based on bonafide financial need Whenever subsidies are not targeted there are potentially wasted resources. The introduction of basic telephone “lifeline" service based on a "need" (income) test is a good example; this has proven to be much more socially efficient than a blanket subsidy for all local service subscribers, many of whom can afford it. As the financial data provided earlier indicate, many small rural telephone companies have very healthy cash flow situations and do not really need subsidies.

Direct government subsidies for rural telecommunications should be discouraged since the investments funded will presumably generate some level of on going subscriber revenue and should therefore always be included in any loan repayment formula, even if the repayment is only a partial one.

Jump to Section:
Contents, 1, 2, 3, 4, 5, 6, (7), 8, Tables, Footnotes, Charts & Figures

Back to the 1996 Rural Telecommunications Workshop Homepage

Please send any comments or questions about this site to