Interconnection and pricing are central considerations that received substantial attention during the seminars. Accordingly, they are treated specifically in this Section of this report.57
INTERCONNECTION
Why is interconnection so important?
As noted previously, a basic premise of communications is that one measure of a network's value is its ability to interconnect the maximum number of users. A competitive market that produces many small, detached networks is undesirable, relatively useless, and essentially counter productive. Users value the ability to reach all other communicators through a single, simple access point, such as the telephone. Thus, interconnection is an indispensable feature of the desired communications network structure.
To the extent that any company has an existing substantially installed network, interconnection permits competition to develop in a less costly manner. New entrants permitted to interconnect on reasonable terms and conditions will not have to invest in duplicating existing plant; investment can be devoted instead to expansion and upgrade of facilities and services. Of course, some duplication will be required to accommodate increases in capacity. The cost of installing additional switches (with the associated software) is enormous; therefore, access to the switch is a critical feature of any effort to develop a coherent and competitive market. Transport lines are becoming less expensive, although the costs of installation, in terms of disruption as well as actual installation work, may be substantial in many areas.58
A related point is that interconnection increases efficiency by avoiding wasteful duplication. So long as interconnection charges are cost based, and so long as interconnection arrangements do not pose excessive externalities on new users (such as requiring the users of a new network to dial a substantial number of additional digits to be interconnected), no facilities should be duplicated until they can be constructed at a lower marginal cost. In short, assets will be devoted to their most productive uses.
Is interconnection a precondition for competition?
At the very least, it is a necessary element of a useful, competitive market. The goal of any communications market today is to provide users with a seamless network. Such a goal has both economic and technological components:
The technical complexities of interconnection efforts and the relationship of standard-setting processes to the functioning of a free and open market impose a particularly significant set of difficult concerns on any effort to develop interconnection arrangements. As regulators and policy makers become enmeshed in the evolving technological characteristics of the intelligent network, their efforts to promote competition may run headlong into the commercial and economic considerations of competing companies. For example, when the FCC proposed its open network architecture (ONA) and comparably efficient interconnection (CEI) requirements, discussed in more detail below, a representative of AT&T offered the following concerns:
In issuing these requirements, the FCC's objective was to ensure equal treatment of all enhanced service providers. But through its ONA proposal, the [C]ommission, in fact, has put in place a set of burdensome regulatory requirements that appear to go well beyond the degree of unbundling required to meet the foreseeable needs of enhanced service providers. For example, the basic service capabilities that compose AT&T's network are not designed to function as stand-alone elements. Embedded within the network, these elements openly interact and combine with one another to produce AT&T services . . . .When it reviewed the initial ONAplans submitted by the RBOCs in response to the ONA requirements, the FCC appeared sensitive to these concerns. Responding to criticism that the RBOCs were not proposing a sufficient level of network openness or unbundling, the FCC wrote, "While fundamental unbundling could be a socially desirable goal, we do not believe that it would be in the public interest for this Commission to require such a development at this time. To require the BOCs to provide ONA based on a more disaggregated, and as yet unspecified architecture would be extremely costly and disruptive . . . . [S]uch unbundling could potentially present technical or operational difficulties. . . ."61 These tensions, between policy and technology, and between the proprietary and economic considerations of competing firms and government efforts to promote an open and accessible market, have pervaded interconnection debates over time.To fulfill the [C]ommission's requirements, substantial modifications to AT&T's network would have to be made to produce discrete elements with protections that would ensure the technical integrity, reliability, and security of AT&T's network. At great expense, AT&T would have to construct new walls and interfaces within the network without knowing whether the demand exists or whether enhanced service providers or end users would be willing or able to pay for these unbundled components. . . .
AT&T is in the basic service business, not the basic service functions business. If ONA requirements force AT&T to give away piece parts of its basic network, the company will have little incentive to develop new basic service capabilities. AT&T has no incentive to offer network-based enhanced services if it means exposing proprietary technology to basic service competitors . . . .60
Are CEI and ONA separate pieces of a U.S. policy to accomplish broad network interconnection?
These are elements of a regulatory strategy ultimately intended to provide all users (end users and service providers) with broad access to a range of services at reasonable costs. To reach this result, the FCC has tried to ensure that anyone seeking to provide (or simply to make use of) a particular service is given reasonable access to and use of the features of the increasingly intelligent network that are necessary to the functioning of such service.62 CEI, as its name implies, may be properly viewed as a fairly straightforward requirement of equal treatment through interconnection; but ONA is a more ambitious effort to organize networks along principles of openness and universal accessibility. In this regard, ONA is more than a plan for interconnection; it is a plan for interoperability.63
CEI was adopted as part of a transitional, interim strategy for expanding the openness and accessibility of the intelligent network. ONA is the longer term objective that subsumes CEI. ONA involves a fundamental restructuring of the way in which networks are assembled and offered. A goal of ONA is to ensure that network intelligence is widely available on a nondiscriminatory and highly flexible basis.
CEI focuses on the conditions under which service providers will be given access to and use of the features of a central switch operated by a network owner, such as a telephone company. In essence, CEI requires the switch owner to provide itself and its competitors with roughly equivalent access arrangements, in both economic and technical terms. CEI is designed to ensure that competitors are not disadvantaged -- vis-à-vis the switch owner -- when providing services that depend upon access to the switch and its intelligence.
Under CEI, the switch owner retains control and decides how to exploit its switch. The U.S. government sets no required level of exploitation; it simply requires the switch owner to treat all competitors the way it treats itself. If the switch owner does not exploit the features of its switch to provide new and innovative services, it is not required to let others do so.
ONA is a more expansive, proactive policy. In the United States, ONA consists of a set of policy objectives that require network operators to piece out the intelligence of their switches and make it available to all interested parties on an unbundled and nondiscriminatory basis. If a network is structured and operated consistent with the goals of ONA, the network operator cannot prevent others from exploiting network intelligence.
ONA is a long-term process that will be achieved -- if at all -- only through the evolution of network structures and offerings. Because ONA seeks to change both the principles by which a network is engineered and the terms on which it is made available to users (including both service providers and end users), ONA will affect both network expansion and development plans, as well as tariffed offerings. Eventually, ONA will subsume CEI because it requires by definition that nondiscriminatory access to the switch be provided as a component of the open network.
The FCC's ONA policies require network operators to file detailed ONA plans (which change over time) demonstrating how network intelligence will be made available, in small pieces, to all interested parties.
Who can be expected to pay for these changes, even though they are desirable?
CEI is, and should be viewed as, a transitional safeguard against unfair competition. ONA, by contrast, is an attempt to promote the evolution of the network to a point where the existing investment base in switch technology can be exploited readily by all interested parties.64
As a general matter, users should be expected to pay for them. Investors may finance them, but users ultimately will have to pay through rates that produce an adequate return on investment. If the investor has no reasonable likelihood of earning such a return, the investment probably will not be made.65
Generally, policy makers are challenged in the area of interconnection policy when they try to find the balance between the level of charges the market can absorb (because of the existing level of demand for service) and the level of change needed to promote general economic development, foreign business investment, and various social objectives. Innovative pricing arrangements, including "value pricing" (discussed below), may be useful. Those parts of the market that demand, and can afford, the greatest degree of access and interconnection may be charged prices that will help fund more broad-based changes. Over time, all prices are reduced as investments are recouped, depreciated, and absorbed into the overall operating structure.
Must the government set interconnection policy, or can the market be permitted to determine the terms and conditions of interconnection?
Both the government and the market are likely to play roles in the interconnection process. The market may be able to accomplish many, if not all, of the tasks involved in establishing terms and conditions for interconnection. The government may have only to be available (or to threaten its "availability") in the event that market negotiations fail or yield proposals that new entrants believe are unfair or unworkable.66
In some cases, the government may take a more active role by requiring companies to follow a broad, general framework that reflects a new or more ambitious national policy. ONA in the United States is one example of such a role for the government. Though the FCC is not dictating the technical and economic details of the "open network" to network operators, it is setting out general parameters and reviewing ONA plans to ensure their compliance with them. It also will act, as necessary, to enforce adherence to ONA principles and the terms of the companies' own ONA plans.
PRICING
As noted previously, one of the desirable features of a competitive market is that it virtually eliminates the need for government involvement in pricing decisions. Because any regulatory process is inevitably imprecise, slow to respond to market developments, and strongly influenced by social and political forces, regulatory decisions about prices are imperfect. Prices established by a functioning market mechanism, in response to the pressures of supply and demand, are likely to be far more satisfactory.
Although ending or minimizing government involvement in pricing decisions is desirable, the government will probably have to be closely involved in these decisions for some time, partly because restructuring is likely to require some rebalancing of rates among different services. Rate rebalancing is politically sensitive, and the government is certain to be involved, whether or not such involvement is desirable.66A In addition, as a market moves from a noncompetitive to a more competitive state, there may be a need for price scrutiny. Such scrutiny may be required to ensure that a dominant firm does not use its market power in one sector of the market to establish anticompetitive rates in another sector. And since a former monopoly telephone company is likely to retain some degree of market power for a while, at least in some sectors of the telecommunications market, price regulation to prevent monopoly pricing strategies may be necessary or appropriate.
With these qualifications in mind, the following discussion considers various aspects of government involvement in pricing decisions.
To evaluate a proposed system of price regulation or oversight, it seems necessary to determine the role that prices are intended to play. Are there different roles for prices and, therefore, different objectives of pricing regulation?
Economic regulation must take account of the various potential objectives of a pricing system. For example, prices might be set to accomplish any of the following:
How does the FCC evaluate rates?
The Communications Act requires the FCC to establish rates that are "just, reasonable, and nondiscriminatory." By law, the FCC has four basic options available to it to ensure that rates submitted to it meet these standards:
Historically, the telephone companies have submitted their rates to the FCC, and state regulators as well, in the form of a tariff. A tariff is nothing more than a kind of contract that sets out all of the terms and conditions of service -- recurring charges, non-recurring charges, and the characteristics and availability of service. Because the tariff is developed to inform users of what they can expect to receive and what they can expect to be charged, it is available for examination by the public.
If the FCC decides to investigate a tariff, it can require the service provider to submit extensive cost and pricing data; the service provider has the burden of justifying the rate as proposed.
As explained below, this process has changed in the United States with the movement toward price caps. In spite of the change in process, however, the ultimate objective, as set forth by the Communications Act, remains the same: the FCC is still responsible for ensuring that rates are "just, reasonable, and nondiscriminatory."
The price cap system seeks to eliminate scrutiny of particular rates by permitting service providers to charge whatever they want to -- within a particular range of a specified cap -- so long as the aggregate price charged for the basket of services subject to the particular cap does not exceed a specified percentage above or below the cap.68
The discussion that follows examines both price caps and the older practice of evaluating service tariffs under a rate-of-return regulatory scheme.
Price caps were adopted in an effort to move away from profit regulation. Why was this shift advocated, and how, briefly, does a price cap work?
There is great disagreement over the focus of economic regulation, and the answer depends in part upon the objectives of the regulation.
An increasingly popular view is that regulation should focus on prices, not profits. Under this theory, if prices are at a level that attracts users, investment will flow into the market. Competition will ensue, which will keep profits at a reasonable level by exerting a constant downward pressure on prices. Advocates who evaluate the performance of a market by focusing on prices argue that the price-demand relationship is far more important than the cost-profit relationship and that service prices should reflect costs and demand. Under traditional profit regulation, such as rate-of-return regulation, prices could be set at any level, so long as aggregate revenues equaled investment, increased by a stated profit percentage. Because the supplier typically was a monopolist or oligopolist, prices did not have to take any real account of demand.
Determining an appropriate level for prices is a problem under either price or profit regulation. Where the market is not competitive, or not readily competitive, the government may have to monitor prices to guard against gouging and other anticompetitive practices. In that situation, a regulator may turn to price caps.69 Because the initial cap usually bears some relationship to the price that existed before the cap was imposed, it affords some protection against increases, with the exception of inflation and similar nondiscretionary factors. If the regulatory system is designed to encourage greater efficiency, or rests on the assumption that prices are artificially high and should be forced down over time, the cap may provide for adjustments, perhaps by using a productivity index or other factor to limit annual increases. These regulatory devices effectively require prices in the telecommunications sector to rise more slowly than prices in other industry sectors.
Consider the following example of a price cap, which is based upon the approach adopted by the FCC.70 Assume that (1) a price cap applies to two services of equal demand (Service A and Service B), (2) a change within 10% of the cap is permissible, and (3) the cap is determined by the following formula:
The Price Index is a unit of measurement that reflects the impact of inflation, either in the economy generally or in some appropriate market sector (such as information products). The Productivity Factor, which is set by the regulator, is reviewed periodically. If the current price equals 100%, the GNP increase equals 5%, the Price Index equals 3.3%, and the Productivity Factor equals 3%, the price cap would be calculated as follows:
If it is proposed to increase the price of Service A from $1.00 to $1.10, and to reduce the price of Service B from $1.00 to $.80, the regulator must determine whether the new price structure falls within the cap. To make this determination, the regulator tests the changes against the cap, in each case multiplying the service's price change by its relative demand. Because each service has equal demand, the price change for each service is multiplied by 50%:
Because 95% is within 10% of the 98.7% cap, the regulator concludes that the proposed changes are permissible.71
As adopted in the United States, caps are applied to groups -- or "baskets" -- of services, instead of individual services. For example, AT&T's price caps apply to three baskets: residential and small business services, "800" services, and other services. The general composition of each basket reflects concerns about social and economic features of the services in the basket and enables the FCC to implement various policy objectives, such as protecting smaller users from significant price increases. The first basket includes services of great social and political concern. The second basket includes services as to which the FCC concluded that AT&T retains substantial market power. The last basket includes the services where AT&T faces the stiffest competition. In addition to review of the aggregate price of services in each basket, the FCC's regulatory scheme envisions that the prices established for services in each of several categories within a basket may not increase by more than a specified percentage per year, even when the aggregate increase for the basket is within the cap. Thus, for example, to protect residential users, the FCC prohibited AT&T from increasing rates on evening and night/weekend measured toll service by more than 4% per year.72
One danger of price-based regulation is that companies may seek additional profits by reducing investment, permitting service quality to deteriorate, and imposing other cost-saving measures undesirable from a public policy standpoint. Profit-based regulation, such as rate-of-return regulation, was designed in part to guard against this risk; by ensuring that the operator could charge prices that would produce a set level of return on all investments, rate-of-return regulation made investment attractive. Through this regulatory scheme, the government could guarantee that the investment would be profitable (although not too profitable). However, this approach had some unfortunate side effects: it arguably encouraged overinvestment and prices that did not accurately reflect user demand, service value, or truly essential operating costs.
Because of these concerns about investment and overall service quality, which will be particularly acute in developing countries, a sensible scheme of price regulation must have some mechanism for encouraging or rewarding investment and high-quality service. If price regulation is imposed on one firm (or a small number of firms) operating in a market, the market itself may be the stimulus.73 If the market is open and users demand and will pay for high-quality, advanced services, the market will ensure that they are provided.
Alternatively, the government may want to promote longer-term investment or to offset perceived "market failures" by requiring firms to meet certain specified levels of investment and service quality or face administrative sanctions. Possible strategies are discussed subsequently.
In contrast to price caps, what are the fundamentals of rate-of-return regulation?
Even the fundamentals of rate-of-return regulation are complex, because this system of economic regulation itself is complex. However, several important concepts and objectives can be briefly explained.
Rate-of-return regulation assumes that the regulated entity is either a monopoly or a company facing little competition; therefore, it behaves in a way that would not be possible in a competitive market. By restricting profit (the "rate of return" on the regulated company's invested assets), this regulatory approach tries to prevent the regulated company from earning excess profits.74
In addition, rate-of-return regulation, as implemented in the United States, promotes investment and expansion of the network. Because the U.S. regulatory scheme permits the regulated telephone company to charge prices that will earn it a stated return on all invested assets, the company has some incentive to expand and modernize the network. In fact, one of the criticisms of rate-of-return regulation is that it overemphasizes investment: when the telephone company earns a stated return on investment, it may overinvest; thus, it may spend funds for assets that it would not acquire in a competitive market, where market prices would not permit it to earn the kind of return it can earn in the noncompetitive, regulated market. With the rate of return permitted by the regulatory scheme, the telephone company can make an investment decision without having to face the normal pressures of attempting to match supply, demand, and prices.
This form of regulation also has been criticized as promoting inefficiency, or not sufficiently encouraging efficiency. Because the telephone company can earn no more than the stated rate of return, it has only limited economic incentives to cut costs and boost efficiencies.75 After it has reached its maximum profit level, it has little incentive to seek greater profitability through cost cutting and efficiency because additional profits would have to be returned to rate payers in the form of a refund. In fact, the telephone company has an incentive to increase costs to the point at which it earns no more than the maximum profit percentage.
On the positive side, rate-of-return regulation in the United States has created a national telephone system that provides reliable, modern, reasonably affordable service to virtually all citizens. Arguably, it also has exerted pressure on monopoly telephone companies to build out their networks and upgrade their equipment.76
These basic concepts are not difficult to understand. The difficulty arises in the implementation of this regulatory scheme. Because the basis of the regulation is costs, the regulator and the regulated company must be able to identify and allocate costs. As practiced in the United States, rate-of-return regulation has been applied to different classes of service. However, the cost base for switched residential service, for example, is different from the cost base for business private line service. Unfortunately, there are many common and shared costs that cannot be allocated precisely -- or even logically -- to particular services or groups of services. As a result, companies and regulators must estimate (skeptics would say "guess at") cost allocations. In the absence of any clear allocation principle, their estimates can be completely arbitrary. In many cases, regulators bargain with the network operator to develop the cost allocations.
The difficulty of cost allocation is both a principal weakness and a legitimate source of criticism of rate-of-return regulation. If cost estimates are too high, revenues are artificially inflated. If they are too low, revenues are artificially suppressed.
Eventually, these cost allocations are used to develop a cost base referred to as "net plant plus overhead." This is the cost base on which U.S. regulators must calculate the "revenue requirement" for the regulated company. The revenue requirement is the amount of money that the company needs to earn to cover its costs (net plant plus overhead) and to produce a profit that does not exceed the maximum permitted percentage. Of course, to arrive at this revenue requirement, regulators must decide on the permitted profit percentage -- the permitted rate of return.77 This rate of return is based upon the regulator's estimate of the return on investment that a company in the industry should be permitted to earn, in view of the return earned by companies in similar businesses. Of course, this, too, is a negotiated process.
In short, allocated costs plus the rate of return equal the revenue requirement.
After the revenue requirement has been determined, the regulated company may set rates to produce revenues that will not exceed the revenue requirement. Any excess revenues must be returned to rate payers in the form of rebates.
Finally, the other major disadvantages of rate-of-return regulation are that it is both expensive to administer and extremely time consuming. The regulated company needs a sophisticated cost and accounting system.78 It also will be required to devote substantial economic and personnel resources to the cost allocation process and to working with the regulator, who also must have substantial accounting and economic expertise. This process has added enormous administrative and regulatory costs to the cost of providing telecommunications service. In a more competitive market, or a market that uses a less onerous form of economic regulation, these costs would decline: the regulator and the regulated company would be able to reduce overhead expenses to benefit rate payers, as well as the public generally (to the extent that the regulator's budget consumes a smaller amount of public funds).
If rate-of-return regulation tends to promote investment and does not typically produce service quality reductions by companies looking for additional profit, it seems that a competitive market, or price cap regulation, might result in underinvestment and excessive cost-cutting that reduces quality of service. What can be done to prevent this?
If rate-of-return regulation encourages "gold plating," price cap regulation may be criticized for encouraging "tin plating," as noted previously. A competitive market can create substantial incentives to cut costs, which may reduce investment and service quality.79 However, actions can be taken in each case to forestall such results.
First, as previously noted, price cap regulation can include a productivity factor to keep periodic increases in the cap under the inflation rate; if the factor is high enough and inflation is low enough, prices may experience a forced decline. The problem, of course, is that it is difficult for a regulator to determine an appropriate productivity factor.80 In the United States, the factor is based upon historic productivity increases in the industry, as well as productivity gains in similar industries. However, if the productivity factor is too ambitious, firms will cease to be profitable, leave the market, or try to cut costs in undesirable ways (for example, by delaying or eliminating maintenance). An unreasonably small productivity factor will permit companies to earn desirable profit levels with prices that simply keep pace with inflation but do not reflect increased efficiencies.
Second, the government might strike a deal with a regulated company that involves some examination of costs and profits. For example, the government might concurrently regulate prices with a price cap and work out an agreement with the telephone company that profits in excess of a certain percentage would be allocated between stockholders and rate payers, according to fixed percentages. For example, a regulated company earning a 12% return might negotiate an arrangement that would permit it to retain 50% of all profit from annual returns on investment of more than 12% and less than 15%. Under the agreement, the company's share of profits over 15% would decline until it reached 0% at a stated higher level of profitability. Such an arrangement would give the company an incentive to reduce costs, thereby boosting profits, while operating within the price cap. Although this also would encourage efficiency, it could produce some undesirable cost-cutting.
Another disadvantage of this approach is that it requires some continuing attention to cost allocation. Price caps, of course, are designed to help regulators avoid the difficult, inevitably arbitrary, task of monitoring and allocating costs.
This type of agreement would be enhanced by a provision requiring the company to reserve a specified portion of profits retained as working capital or otherwise to hold such funds as undistributed retained earnings. The balance of profits retained by the company would be free for distribution to stockholders as dividends.
In the United States, these kinds of sharing agreements have been developed with regulated telephone companies as prerequisites to reducing the extent of general regulation imposed upon such companies. Loosely described as affording a "social contract" approach to regulation, these agreements make the government and the regulated company parties to arrangements under which users of the networks receive a significant portion of the increased profitability that less regulated telephone companies may enjoy.81
Third, the government might adopt certain minimum engineering standards. If they were combined with some type of enforcement mechanism, these standards could be used to ensure that telephone companies were not increasing profitability by reducing investment in maintenance, repair, and reasonable upgrades.
Fourth, the government might develop some type of program for accepting and investigating consumer complaints, in order to factor such complaints into other regulatory decisions, such as licensing or price regulation. The government also could require the regulated telephone company to provide periodic reports addressing (1) the number of complaints received during a particular period, (2) the increase or decrease of such complaints, and (3) the actions taken by the company to resolve complaints. Enforcement mechanisms can impose penalties -- including monetary fines -- on companies with excessive rates of unresolved complaints.82
All of these mechanisms are somewhat problematic because they require the government to make judgments about the level and quality of service that are appropriate for a particular market. This means that the government must make economic decisions in light of supply, demand, cost, and revenue information. Even when public process encourages extensive participation, as in the United States, it is very difficult for the government to obtain sufficient information to enable it to make such decisions with any precision. For this reason, many critics of regulation argue that only a competitive market can make truly efficient decisions about investment and service quality. In a competitive market, firms must respond to consumer demand for upgrades, better service, faster repairs, and similar items; and they can make economic decisions about such matters, based on the costs of providing the requested facilities and services and the willingness of potential users to pay for them. At the very least, this approach ensures the efficient, cost-effective use of available resources.
In some markets, a perplexing problem arises when a low level of competition or consumers' inability to pay (in a poor urban neighborhood, for example), results in service levels that are considered unacceptable, as a matter of social policy. In such instances, it may be appropriate for the government to impose a minimum service requirement. When developing this requirement, however, the government should be careful to determine a level of service that realistically can be funded, as opposed to the level of service that might be desirable without regard to economics. To avoid producing significant distortion in the marketplace generally, the government's minimum service requirements should be coupled with some kind of subsidy mechanism to ensure that the service provider receives at least the costs of providing the minimum required level of service. Without this, firms will either shun the market or begin to develop hidden, internal cross subsidies that will ultimately produce widespread, undesirable marketplace effects.
By limiting the extent of its involvement in this type of activity, the government can avoid imposing inadvertent ceilings on development. More advanced services should evolve freely in response to market demands.83 Instead of prescribing general levels for the provision of services, the government should concentrate on setting absolute minimums, which may mitigate the probability that poverty will create a sector of society that is both information poor and communications deprived.
"Value pricing" was mentioned earlier (in note 67) as a way of using prices as a rationing tool. Is such an approach inappropriate for an essential service such as telephony?
Such a use of the pricing mechanism might seem objectionable because of social considerations. Denying individuals access to basic communications service (even temporarily) because of their economic condition raises grave concerns about fundamental fairness. As a practical matter, it also sparks substantial political resistance. However, the following unattractive alternatives to "value pricing" should be considered:
In each of these cases except the third, price reductions intended to guarantee wider access do not appear to make great sense. Nor do they seem likely to accomplish their objectives. Increased access is a very desirable goal, but it is not necessarily true that prices are the key to ensuring it. In fact, prices might more properly be viewed as a mechanism for promoting profitability that will foster expansion and upgrade, eventually resulting in a more universally affordable network arrangement. If the government does not want to deny users access to the network when both development and operating costs are high, it must decide whether to subsidize development by the network, payment by the users, or both; it cannot simply ignore the relationships among prices, revenues, profits, reinvestment, efficiency, and economies of scope and scale and insist that an inefficient, poorly developed network should operate like an efficient, highly developed one.
The benefit of value pricing is that it permits prices to reflect not only the costs, but also the realities of supply and demand. If the price that results from this approach is considered too high for social reasons, the government may want to consider providing some users with subsidies, instead of arbitrarily requiring prices for some services to be set at a lower level (offset by artificially inflated prices on other services). To avoid excessive fiscal demands on public funds, such subsidies should be targeted as precisely as possible to reach those who need them most. Such targeting may be facilitated by considering such policy matters as the definition of "universal service," making every effort to identify the minimum feasible level of service to be provided, in light of technical and economic constraints. In general, subsidies should be aimed at two large groups: those with the lowest incomes and those living in areas that impose the highest costs on service provision.
Interconnection and pricing policies are among the most important and complex aspects of telecommunications regulation; and they play critical roles in any effort to promote a more competitive market environment. Without interconnection, fledgling competition almost certainly will languish. Without clear and sensible pricing policies, market development, social considerations, and efforts to encourage entry are likely to be confused and unsuccessful, as well as targets for political dissatisfaction.