RESPONSE TO THE NTIA REQUEST FOR INFORMATION ON BROADBAND
William J. Baumol*
Professor of Economics, New York University
Senior Research Economist, Princeton University
Telecommunications is an industry that has been in the forefront of innovation in the United States and elsewhere. It continues to play a vital role in many areas of national technical progress, one of the most important being broadband access. Recent discussions over the role of government in encouraging such advance have raised questions about the current telecommunications laws and regulations, and their implications for investment in broadband facilities. In particular, several parties have suggested that the pricing and unbundling policies adopted by the Federal Communications Commission (“FCC”) under the Telecommunications Act of 1996 are in need of modification because the resulting unbundling requirements and limitations on the charges permitted to the incumbent local exchange carriers (“ILECs”) serve as a disincentive to further investment in the “last mile” facilities necessary to provide broadband access to the majority of U.S. customers. It is argued, moreover, that what is appropriate is further deregulation of the local telecommunications arena, without preconditions, in order to permit the industry participants to be guided by the unrestricted forces of market mechanisms rather than government fiat.
In light of the importance of these considerations the NTIA has issued a request for comments on the development of broadband networks and advanced telecommunications, and has provided a set of questions to guide these comments. In this memorandum I offer comments on the several of these questions that I consider myself competent to discuss.
Before turning to particular questions from the NTIA Request, it is important to lay out several general principles that are fundamental to understanding of the economics of broadband access. In particular, those who support substantial revision of the current unbundling rules and who criticize reliance upon total element long-run incremental cost (“TELRIC”) as the cost concept appropriate for the pricing of unbundled network elements (“UNEs”) appear to rely on two premises: (1) the notion that markets always serve the public interest more effectively than regulation by the public sector; and (2) that a restriction upon earnings inherently constitutes an undesirable disincentive for investment in the affected arena. Both of these proffered standards are related to principles that are entirely defensible, but only when applied to fully competitive markets – not markets such as those in which ILECs today provide access to broadband UNEs and services. Indeed, blind application of each of these principles to today’s broadband access and UNE markets would lead to policies that if adopted would be severely detrimental to the general interest.
Fallacy I: Deregulation of Monopolized Markets. First, it is critical to recognize that it is not markets per se, but effectively competitive markets that can be relied upon, in the absence of regulatory constraints, to enhance output and investment and to serve the public most effectively. It is, of course, widely recognized and confirmed by evidence and analysis that effectively competitive markets serve the public interest by promoting innovation, by bringing investment and output to maximal levels consistent with balancing of cost and revenues, and by charging prices as low as financial viability permits. But matters are very different in markets dominated by firms that possess monopoly power. One of the well recognized attributes of a firm with monopoly power is its restriction of investment and output. The very definition of monopoly power is the ability to charge prices substantially higher than competitive levels, and to retain those high prices for a significant period of time. But the way in which such a firm is able to extract high prices from consumers is by the creation of artificial scarcity – by restricting outputs and the investments used to supply those outputs, to quantities well below those that would prevail under competition. The firm’s incentive to restrict investment in such markets is further enhanced where such investment can be expected to render current plant and equipment obsolete. There, the dominant incumbent, immune from material competitive pressures, can be expected to resist such change by keeping its investments to a minimum, protecting the earning power of its old equipment until and if management’s hand is forced by the incursion of substantial rivals whose more modern facilities threaten the business of the incumbent.
Thus, in markets dominated by such firms, there are only two ways in which the public interest can be protected: by finding ways to introduce effective competition, or by regulatory constraints that prevent the exercise of monopoly power. In some cases, arguably in the local telephone sector, a combination of the two is the most promising approach.
But it is outright fallacy to expect the public interest to be served by elimination of effective regulatory rules in markets where, without regulatory oversight, the public would be left to the tender mercies of dominant firms with the ability to exercise monopoly power. It is equally fallacious to expect measures that are tantamount to protection of the incumbents from effective competition to result in stimulation of investment and output. That is the reverse of what unregulated monopolists can be relied upon to do.
Fallacy II: The More Reward the Better. There is some truth to the assertion that an increase in earnings on new investment that is made possible by changes in regulation that permit enhanced prices for access or some other item supplied by the regulated firm may elicit enhanced investment. But there is no validity to the suggestion that this must invariably serve the welfare of the consuming public. If the monopolist is enabled to charge prices higher than those previously permitted, it is a truism that this offers a higher return per unit of output and that the financial yield on investment may thereby be stimulated. But there are two things wrong with the succeeding step in the argument. First, why then is the sky not the limit? If the higher the firm’s prices the better, why not encourage prices well above any correspondence to costs? Once stated this way, the fallacy is evident. Consumer interests must be presumed to be promoted by low prices, not by high prices. This is true even if the prices in question are not those charged to consumers, but to competitor firms who supply final product to the public, and whose final product prices must be raised to make up for the high charges these competitors are forced to pay the dominant firm or, who must exit from the market because of inability to recover those high charges.
But, here too, the fallacy goes even deeper. For increased prices, whether to consumers directly or to competing suppliers, have a predictable consequence. High prices are very effective in cutting customer demand. Thus, they will lead to what is evidently a main current impediment to expansion of broadband investment – lack of demand. The evidence appears to show that already broadband capacity has considerably outstripped demand for its output in that over eighty percent of households have DSL or cable modem services available to them, but only about ten percent actually subscribe.
Thus, the notion that investment in broadband will be enhanced by the imposition of UNE prices that exceed those called for by TELRIC is patently indefensible. Higher access prices can be depended upon to keep the volume of customer demand lower than it would otherwise have been, either because they will result in higher customer prices charged both by ILECs and by their competitors; or because they will prevent effective competition in the local arena altogether and that, surely, is not the way to stimulate investment – either from the competitors or from the ILEC.
Having dealt with these preliminaries I turn to responses to particular questions posed in the NTIA’s Request. I shall confine myself to Question A, and to the four sub-inquiries of Question E because these are the matters raised in the Request on which I feel most qualified to comment.
Question A: Primary Considerations for Broadband Policy
Economic analysis offers clear guidelines for the considerations that merit priority as they relate to issues such as broadband policy on pricing, unbundling, output, investment, competition and regulation. Taking as a basic premise that the ultimate goal of such policy should be promotion of consumer welfare, the rest follows. For experience and analysis concur in the conclusion that this goal is most effectively promoted by deregulation, but only if preceded or accompanied by effective competition. Competitive markets yield the prices that serve as the required incentives for efficient levels of investment and output. They enforce productive efficiency along with the low customer prices that the resulting low costs permit. They stimulate the innovation that is so critical for achievement of the benefits that can be expected to derive from broadband.
For competitive markets to emerge in an arena such as broadband, however, it is necessary that there be an efficient market for access to the necessary local telephone facilities. There are two attributes of these telephone facilities that are necessary for broadband to reach most American homes that lead to this requirement for feasibility of effective competition: first, the facilities, many of which are already in place and which serve both “narrowband” and broadband traffic, are very costly; and, second, because of their economies of scale and scope, they have the capacity to serve the broadband access needs of prospective and actual competitors at much lower incremental cost than if they all had to be replicated by an entrant. In these circumstances, it would be highly wasteful, and very costly to the economy for competitors to be required to replicate all currently available facilities. This is a high cost that inevitably would have to be borne by consumers – patently no way to advance their welfare. More than that, if replication of all facilities were the only avenue to entry, it is probable that effective competition would never emerge. A foretaste of this prospect is provided by the glacial pace of substantial competitive entry into “last mile” facilities (outside of dense business districts) in the local telephone arena, and by the exit of many competitors from broadband activity.
Where costly advanced service facilities are required, the only way in which the entry of competitors can be ensured is through an active market in access to the incumbent’s facilities. The key role of regulation here is a treble one: first, to require incumbents to provide competitors access to all network elements that it is inefficient for these competitors to provide for themselves; second, to require that the prices charged for such access be compensatory but not exclusionary – for a sufficiently high access price can patently nullify any access availability arrangement; and third, regulation must also ensure that freedom of access not be undermined by forcing competitors to rent facilities they do not need, but that the incumbent may have bundled with the others.
It should also be emphasized that the sale of such access is not an act of charity forced on the incumbent. On the contrary, even competitive firms in competitive markets find it profitable to sell access, just as other firms voluntarily and profitably sell access to their patented technology, to their competitors. For example, in the interexchange business, firms such as Level 3, Sprint and Williams – without any hope of monopoly returns – are willing to act as carriers’ carriers and sell long distance capacity to their rivals. Clearly, markets for such access can prosper at competitive prices. But it is just as clear that in the absence of regulatory directives, monopoly ILECs will choose not to sell such access, or to sell it only at monopoly prices. However such ILEC behavior tells us nothing about whether the sale of access is profitable at competitive prices – only that it is more profitable at monopoly prices.
Question E1: Do Regulatory Rules Cause Investment Disincentives?
Regulated rates and other statutory requirements upon the incumbent, such as required interconnection, unbundling of network elements and resale of services may possibly serve as a disincentive for investment. But whether they do constitute such a disincentive and whether the disincentive is improper, depends on the nature of the facilities and interconnection to be supplied, and on the magnitudes of their permitted prices.
First, if access to these ILEC facilities and interconnection is necessary for a competitive carrier to offer broadband service to customers economically, and unavailability of such access would impair its service offerings, regulations that require such access to be offered promote efficient investment, they do not impede it. And for that reason, the 1996 Telecommunications Act’s requirement that such necessary access be supplied constitutes no distortion of the proper incentives for investment.
Second, so long as prices are set at the cost-based levels that prevail in competitive markets, investment incentives will be those consistent with the requirements of economic efficiency. But this, too, is what the 1996 Telecommunications act requires. Obviously, there are ways in which higher levels of investment could be elicited, for example by generous government subsidies financed by increased taxes, but that surely is not the way to promote the public interest.
Perhaps even more important, if the rules are changed with the intention of promoting investment in broadband, but in a way that simultaneously preserves monopoly and reduces regulatory oversight, the likely result is less investment, not more. The incentive of a monopolist, as already discussed, is to hold back output and, therefore, to rein in investment correspondingly. In addition, if the investment promises to provide superior facilities that render current capacity obsolete, why should a firm that is insulated from competition and not subject to regulatory control of its investment outlays increase its investments?
Question E2: How to Recover Added Network Costs of Serving Competitors?
In competitive markets prices are based on incremental costs. But those incremental costs must in the long run cover the cost of any investments required to serve the needs of both ILEC and CLEC use. That is, incremental investment required to serve CLECs is not something to be added to long run incremental cost. It is already a basic component of that long run incremental cost figure – if the figure has been calculated properly. The same is true of any associated risks, as will be discussed in my comments on the next question. Thus, there is no need for a separate, additive calculation of the depreciation and risk cost entailed in recovery of incremental investment, unless it can be shown that a state’s long run incremental cost calculation (under the TELRIC standards adopted by the FCC), is, in the case of broadband, not carried out properly.
But there is another relevant consideration. This pertains to the magnitudes entailed in the matter. I have seen little to suggest that the magnitudes of incremental investments needed to serve competitors’ demands are substantial. For the moment, this is suggested by the fact that UNE use of ILEC facilities currently amounts to less than 3 percent of total ILEC lines, meaning that it is hardly plausible that any incremental investments required to serve this demand act as a substantial disincentive to investors. And over a longer run, ILECs claims that they face the threat of stranded costs provide strong reason to doubt that substantial incremental ILEC investments will be required to meet competitor demands. If it is true that significant competitive entry will result in stranded incumbent facilities, then it follows that little investment in added facilities will be needed to serve the entrants. The ILECs cannot have it both ways – entry cannot cause both significant unused capacity in the form of stranded investment and a need for substantial added investment. In any event, UNE entry by CLECs must surely reduce the risk of stranded investment for the ILECs relative to the risk they would face if entry were exclusively facilities based..
Question E3: What of Risk of Investment to Provide Broadband?
Free markets are inherently risky. The provision of service by any competitive firm always entails the risk that anticipated demands will not materialize. That is just as true for the products that a firm sells to final consumers as it is for products it sells to other firms, so there is no justification for exclusive focus on the latter source of risk rather than both. Risk is a real cost, and prices in any risky market must cover that cost in the long run. This does not mean, however, that this risk is an addition to the supplying firm’s incremental cost because, properly calculated, the firm’s incremental cost should already include the cost of this risk.
The regulatory rate making process takes account of such risks in three ways. The first is in depreciation rates. If a firm faces the risk that its network will lose value because of expected technological improvements or competitive pressure, it is appropriate to build into prices depreciation that is faster than what ordinary wear and tear may suggest. The FCC’s pricing regulations have been scrupulous in allowing depreciation to encompass such risks. The second effect of risk is that on a firm’s cost of capital. But again, because the FCC’s pricing regulations allow costs of capital to be determined from current market data (e.g., bond and equity yields) that reflect investors’ perceptions of ILEC risk, and using calculation methods such as DCF and CAPM that properly incorporate these data, there is no failure in regulatory methods that has prevented such risks from being taken fully into account in ILEC interconnection prices. Finally, regulators may postpone adjustment for some risks in the short run, but make up for this in the long run by permitting recovery of stranded costs. Generally, this is not a preferable way to deal with risk, and it certainly is inconsistent with the competitive market model for regulation in the public interest.
Question E4: Is TELRIC the Proper Cost Measure for Access Pricing?
It is noteworthy that the Bell Operating Companies, who are, apparently, the principal critics of TELRIC pricing standards, have in the past repeatedly voiced criticisms of any regulations that prevented them from reducing their prices to incremental cost levels. This clearly implies that those firms themselves believe incremental cost prices to be compensatory, despite any contrary views they may currently express in a wholesale context. The fact is that unless it can be shown that TELRIC prices are improperly calculated, they are the basis for pricing in truly competitive markets.
Nevertheless, there are two objections of principle that have been raised against TELRIC. First, it is said that TELRIC calculations are inappropriate because they are based on efficient costs rather than actual costs. Second, it is said that the calculation of TELRIC depends on information that is controlled by the firms that are regulated on the basis of these calculations. These matters can be dealt with fairly briefly.
Actual versus Efficient Investment Values. Though at first glance it may appear implausible, in efficient markets, the value of an efficient investment and the value of its “actual” counterpart are one and the same. In taking the contrary position, some firms have argued that they are of necessity saddled with technology of earlier vintages, and in any event, no firm’s equipment choices are ever consistently perfect. Hence, they argue, their “real” cost of these investments is sure to be higher than their counterparts in a hypothetical efficient firm.
This argument has an aura of plausibility, but it has no relation to the behavior of markets of reality. Consider a firm that owns a machine, A, that cost it one million dollars at the earlier date when it was acquired; but suppose that equipment suppliers now offer a machine, B, of identical capacity that requires an outlay of only $700,000. If the firm that is the owner of machine A finds itself in need of cash and now attempts to sell its machine, how much can it realistically hope to obtain for it – the million dollar initial outlay or the $700,000 price of the current substitute? The answer is obvious. It follows immediately the actual cash cost to the firm if it decides to retain machine A for itself is no more than the amortization corresponding to $700,000 – and has no relation whatever to amount it paid for the machine initially. The unforgiving forces of the market dependably and invariably revalue any investment costs to the costs of available efficient counterparts of those investments. Those actual costs are the efficient costs, and any attempt to redefine “actual costs” to cover the inefficiency of some component has no relation to reality.
Regulation on the Basis of Data Supplied by the Regulated Firm. Though there is good reason for concern that regulated firms will succumb to temptation to slant the information they supply as the basis for their own regulation, there is no way around the problem if the competitive market model is to be used, as it should, to guide regulation. After all, in any competitive market a firm’s prices are patently affected by its own costs. But these are invariably available initially only to the firm in question. Whatever standards the regulator uses, I know of no observer who advocates pricing that does not take the regulated firm’s costs into consideration. Thus, the problem is hardly a special manifestation of the FCC’s TELRIC approach. No substitute regulatory framework that purports to base itself on the competitive market model can avoid the difficulty. Only careful oversight of the firms’ cost calculations can minimize the problem.
The analysis leads us to an unambiguous conclusion. The public interest will, indeed, be served by further deregulatory steps, but only in arenas in which competition is not merely an aspiration for the future. The combination of monopoly power and deregulation is not a pairing that will benefit consumers. On the contrary, it will work against their interests by raising prices, impeding innovation and decreasing investment. Experience in the wake of the 1996 Telecommunications Act is entirely consistent with these conclusions. Soon after the adoption of the Act, as competition in broadband made its hesitant appearance, investment rose markedly, both from these entering competitors and from the ILECs. But when the entrants dropped out, ILEC investment in broadband declined precipitately. Of course, ILEC monopoly power is not the entire story – worsening market conditions undoubtedly played a part, both in the exit of competitors and the reduction of ILEC investment. But there is every reason to conclude that the two influences – market conditions and monopoly – both played their roles, and handicapped investment more substantially than either would have by itself.
The bottom line is that there seems to be no reason for tampering with the provisions of the Act or with the FCC’s approach to its implementation. What is called for, rather, is more rigorous enforcement of the Act and the FCC’s rules to ensure rapid materialization of competition – rather than empty assertions of loyalty to this ideal, accompanied by unwavering resistance to its substance.
* These remarks have been prepared on behalf of the Advanced TelCom Group, Inc. (ATGI), the Association for Local Telecommunications Services (ALTS), AT&T, the Competitive Telecommunications Association (CompTel) and WorldCom.
 See “The Cable Industry: Winning the Battle for Consumer Video, Data and Voice, Industry Analysis.” Report by Equity Research Division of J.P. Morgan Securities, Inc., (Nov. 2001), pp. 32, 37.
 See FCC First Report and Order, CC Docket No. 96-98 (In the Matter of Implementation of the Local Competition Provisions in the Telecommunications Act of 1996), ¶¶ 630-740, released 8/8/1996.
 In a state context, see California Public Utilities Commission Investigation I.87-11-033. In the Matter of Alternative Regulatory Frameworks for the Local Exchange Carriers: Decisions D.98-10-031 (p. 144) and D.94-09-065 (pp. 228-229). In a federal context, see FCC Report and Order and Second Further Notice of Proposed Rulemaking, CC Docket No. 87-313 (In the Matter of Policy and Rules Concerning Rates for Dominant Carriers), ¶¶ 502-534 and 806-826, released 4/17/1989.
Of course if it had been anticipated that newer technology would reduce to value of the machine from $1 million to $700,000, this should have been reflected in the depreciation rate assigned to the machine, and thus, would have elevated the initial regulatory price at which the machine’s output could be sold.