U.S. Comments on ICC Framework for Interconnection Charges

Comments of the Government of the United States on the Information and Communication Council's July 27 Report: "A Framework for the Calculation of Interconnection Charges in FY 2005 and Thereafter"

August 27, 2004

The Government of the United States welcomes the opportunity to submit the following comments to Japan's Ministry of Public Management, Home Affairs, Posts and Telecommunications (MPHPT) in response to the Information and Communications Council (the Council) July 27 report on interconnection charges.

The issues addressed in the July 27 Council report that form the basis of possible recommendations cover a range of issues. These comments address three of those issues: (1) proposals to eliminate non-traffic sensitive (NTS) costs from per-minute charges; (2) analysis of NTT's ability to absorb NTS costs through basic monthly charges; and (3) proposal to retain cross-subsidies between NTT East and West, funded by uniform interconnection rates.

(1) Elimination of NTS costs from per-minute interconnection charges

The United States has long argued that rules permitting incumbent carriers to recover NTS network costs through per-minute interconnection charges can lead to inefficient network usage, diminished economic welfare, and an inappropriate burden on competing carriers saddled with paying such charges. Evidence that up to one half (or more) of NTT East and West interconnection charges can be attributed to NTS costs makes this a critical issue for the competitiveness and efficiency of Japan's telecommunications network. While mobile services and VOIP (Voice Over Internet Protocol) have made great inroads in providing competitive alternatives to high-priced origination services, NTT East and West maintain an effective monopoly on terminating calls to the 60 million subscribers they maintain, service which still accounts for the bulk of Japan's total telecommunications traffic. NTT East and West can be expected to maintain their dominance in this sector for the foreseeable future. The recent decline in NTT East and West network traffic, combined with inclusion of NTS costs per-minute interconnection rates, therefore seriously threatens local, long-distance, and VOIP providers' ability to continue offering competitively-priced alternatives. This is because NTT has argued interconnection rates should rise to compensate for the diminished volume of traffic, over which it argues NTS costs should be recovered. Thus, eliminating NTS costs from per-minute interconnection charges is urgently needed to ensure a competitive supply model.

The United States welcomes the Council report statement that "in principle NTS costs should be included in basic charges" (Part 1, Chapter 3) and that "it is necessary to exclude NTS costs from the costs for interconnection charges" (Part 2, chapter 1). The United States has been recommending such steps since 2000, and is pleased that the Council has finally, unequivocally embraced this concept.

Nevertheless, MPHPT should reject the Council report suggestion that this action should be phased in over four to five years. (Chapter 3: "four to five years has been deemed an appropriate time frame for the gradual elimination of the NTS costs from the costs of interconnection charges."). This would inevitably result in an unacceptable increase in interconnection rates in Japan (following increases approved for FY 2004), and put Japan well behind other OECD countries in effectively regulating such rates. Furthermore, it appears contradictory that the proposed phase-in period is longer than the period of applicability of the revised model (three years), with the future calculation method uncertain. Given that a longer phase-in period would clearly be detrimental to NTT's competitors, the study group has not provided adequate justification why a long period would be necessary, nor why an immediate exclusion of NTS costs is not a viable option. Accordingly, adopting this recommendation would fail the standard of impartiality. As discussed below, NTT East and West have been unable to demonstrate that they do not already recover these costs through their monthly subscriber line fees.

(2) Analyzing NTT's ability to absorb NTS costs through basic monthly subscriber line charges

The Council's admission that it does not know the cost allocation associated with NTT's monthly subscriber charges is the strongest argument for denying NTT an extended period to shift NTS costs to this revenue source. The burden of proof justifying any phase-in period should be on NTT. Unless NTT East and West can provide convincing data to the public that their monthly charges cannot cover NTS costs necessary for interconnection, these costs should be immediately eliminated from the LRIC model and considered recovered from these monthly subscriber charges, without raising monthly subscriber charges. It is noteworthy that NTT East and West monthly charges are already relatively high, in comparison with other countries. For example, based on the OECD survey of fixed charges for residential services in 30 OECD countries in 2002, Japan rated as the 7th most expensive. (1)

To adequately analyze the ability of NTT East and West to absorb NTS charges from monthly subscriber charges, several issues must be addressed relating to the transparency of cost accounting of these basic charges, as noted below.

Allocation of initial subscription fee:

Over the years, NTT has accumulated over 4 trillion yen ($30 billion) (2) in initial subscription fees , currently 72,000 yen per subscriber. Based on NTT's own statements, this fee is intended to defray NTT's network construction costs. Whether or not this fee continues to be levied, NTT should be required to fully account for the costs already recovered through this charge, including NTS costs currently recovered through interconnection charges. For example, NTS charges currently allocated to interconnection charges are intended to cover local loop network components such as remote terminals and transmission equipment both of which perform NTS functions. If initial subscription fees have been allocated for the recovery of such equipment, further cost recovery would be redundant. In actuality, NTT may never have transparently allocated initial subscriber fees to specific network elements. Nevertheless, NTT East and West should be required to do so, particularly if they claim that current rates are insufficient to fully recover NTS costs in order to justify recovering these costs from their competitors through interconnection charges.

Allocation of costs for ISDN, DSL, and FTTH:

Over the past two decades, NTT East and West (and their predecessor entities) have invested trillions of yen in network upgrades for ISDN, DSL, and fiber-to-the-home (FTTH). (3) While NTT appears to have curtailed ISDN investment, investment in DSL and FTTH continues. Given NTT's historically aggressive investment strategy in all these areas, the long timeframe required for earning profits from such technologies (and likelihood that ISDN, with its declining subscribership may never be profitable, resulting in significant "stranded investment"), it is not surprising that NTT East and West have, until recently, appeared financially weak and even shown losses. It is unlikely, however, that such losses, result from insufficient recovery of costs relating to traditional telephone services (costs for which interconnection charges are intended to cover), but rather reflect NTT's difficulty in subsidizing the development of new services with its core telephone service revenues.

One reason NTT East and West may be unwilling to offer transparent data on the allocation of the monthly basic charges is that they do not want to divulge the degree to which traditional revenue streams are being used to subsidize the rollout of these new services. (4) Without such clarity, however, the scope of basic service revenue that can be feasibly allocated to interconnection-related NTS costs cannot be identified. Needless to say, it would be inappropriate for NTT to claim that it could not absorb interconnection-related NTS costs from basic monthly charges if its reason were that such revenue was dedicated to funding the roll-out of DSL or FTTH.

This is not to say that NTT East and West are unjustified in using core telephone service revenue to invest in new services. Rather, the United States believes that from the standpoint of competition policy, NTT's first obligation with respect to monthly telephone charges should be to cover the costs (including NTS costs) for the service its telephone subscribers directly pay for (dial tone connectivity to local and long distance networks), and not to foist such costs onto competing carriers.

Another key requirement for a transparent evaluation of revenue streams available to cover interconnection-related NTS costs is to review whether costs shared between different services are being fairly allocated between such services. As is well documented in regulatory theory, regulated monopolies have a strong incentive to maximize costs allocated to their monopoly services, in order to facilitate their ability to capture market share in more competitive sectors. (5) Thus, NTT has an incentive to allocate costs for shared infrastructure (e.g. poles, ducts and conduits, transmission equipment, buildings etc) and personnel (maintenance, sales and marketing, R&D) to its basic telephone services, even where those same resources may be used for alternative services, such as ISDN, DSL or FTTH services. Until NTT East and West accurately document how such shared resources are allocated, the legitimate revenue available to cover NTS costs will be unverifiable.

Depreciation allowances for costs covered by monthly usage fees:

In previous public comments on the LRIC model, many commentors have disagreed with MPHPT's chosen depreciation rates. While the revised LRIC model included more realistic depreciation lives for some equipment (e.g. switches, fiber, conduits) (6) MPHPT appears to permit entirely different depreciation rates for the local loop, which is not regulated based on LRIC principles. (Presumably, covering local loop costs is the main purpose of the monthly subscriber charge.) This raises the same potential concern raised in previous reviews of the LRIC model: if NTT East and West are being permitted to depreciate their local loop costs according to methodologies and asset lives prescribed for tax accounting (e.g. accelerated depreciation), such reported costs may significantly exceed actual economic costs, which is the more appropriate standard for rate regulation. The consequence of overstating such costs is that NTT East and West may be understating the revenue available to cover NTS costs.

(3) Inappropriateness of permitting NTT East to subsidize NTT West through interconnection revenue

As the United States has stated in the past, it is inappropriate for NTT East to be subsidizing NTT West's operations from above-cost revenue obtained from competing carriers. MPHPT should reject the draft report suggestion that this practice be permitted to continue. Permitting NTT East to set interconnection rates above its costs, as clearly identified by the LRIC model, raises questions about whether Japan is acting consistently with its WTO commitment to ensure cost-oriented interconnection. It appears evident, in the case of NTT East, that the interconnection rates are not cost-oriented.

MPHPT has argued that the effect of the fund transfer between NTT East and West is competitively neutral, since NTT East does not keep the above-cost revenue. However, this assertion is debatable. Although NTT East estimates that it will be required to cross-subsidize NTT West to a tune of 19 billion yen in FY 2005, (7) the mechanism for this transfer (above-cost interconnection rates) also shields NTT East's retail service from effective competition by raising rivals' costs, and enables NTT East to price its services at supra-competitive levels. The extra margin NTT East can earn from its dominant share of intra-prefectural traffic (8) likely exceeds the 19 billion yen it is required to transfer to NTT West, to the disadvantage of potential rivals. Permitting such a policy in an environment of declining fixed line traffic, where NTT East is losing market share to mobile services is inconsistent with MPHPT's mission to promote competition in the telecom market to benefit consumers.

Uniform interconnection rates and universal service:

The Information Council report states that "it cannot be said at this stage that a social consensus has been fully obtained for setting interconnection charges separately for East and West, which may lead to a regional disparity in local calling charges that apply to universal service" (Chapter 4). As the United States has noted in previous comments, this logic makes little sense: an increase in interconnection rates in NTT West's territory would raise the cost to NTT West's competitors, but would have no effect on NTT West's internal costs, and thus its ability to maintain current retail rates. Whether NTT West would have enough revenue to cover its costs, absent a subsidy from NTT East, is a separate issue that ought to be examined on its merits - based on NTT West filing the appropriate cost data. Given NTT West's 90 billion yen profit for the year ending March 2004 (9), it is unlikely that an increase in user rates could be justified. Nevertheless, if a deficit were truly identified, NTT West has at its disposal a vehicle to obtain funding without resorting to a user rate increase: the universal service fund that is currently authorized. NTT West's reluctance to avail itself of this funding, based on a desire to keep its cost information confidential, is not compelling. Rather, it increases the likelihood that the cross-subsidy was never justified in the first place.

Need for an analysis of potential price-squeeze by NTT East and West:

MPHPT's willingness to authorize interconnection price increases last year, and consider them again this year raise the serious issue of exclusionary or predatory price squeezes: if competitors' inputs are increased based on an alleged increase in costs, but NTT East and West are not required to reflect such cost increases in their own retail prices, it is likely that competitors will have to either exit the market or increase prices and lose market share. Before even considering any such increase in interconnection rates, MPHPT should, at a minimum, conduct and publicly report the results of imputation tests analyzing the relationship between wholesale and retail rates for NTT East and West.


The United States appreciates the opportunity to provide comments on the Council's report. The proposals contained in the report represent significant progress towards resolving one of the most fundamental issues with Japan's LRIC model: the need to exclude NTS costs. However, the United States urges MPHPT to take more aggressive action to exclude NTS costs as soon as possible and to adjust depreciation rates realistically. The United States also urges MPHPT to reconsider the policy of uniform interconnection rates for NTT East and West, which is anticompetitive and is not necessary to ensure universal service. As MPHPT reviews its policies on interconnection rates, the United States urges MPHPT to seriously weigh these comments and ensure the interests of all stakeholders are carefully and fairly balanced.

1 OECD Communications Outlook 2003, p. 179, Table 6.10.

2 "Like NTT phone fee, line brokers face extinction," Japan Times, July 27, 2004.

3 See, for example, 17 November, 1998 Dresdner Kleinwort Benson research report on NTT, estimating that by FY 2001, NTT was expected to have invested 890 billion yen in ISDN technology (p. 19).

4 Evidence that NTT currently subsidizes advanced services appears uncontroversial: given NTTfs own assertions on its high cost-structure and its history of recent losses, it is difficult to see how 100 Mb/s FTTH service could possibly be profitable for NTT East and West at the current 3,900 yen per month level offered (see NTT East website).

5 See Laffont and Tirole, Competition in Telecommunications, MIT, 2000 p. 145.

6 A key exception is transmission equipment, which, remaining at 8.8 years, appears unrealistically short. The FCC currently prescribes rates of between 11 and 13 years for such equipment. See "Depreciation Ranges Adopted in CC Docket No. 98-137" - Dec. 17, 1999.

7 See "Outline of NTT East's Business Operation Plan for Fiscal Year Ending March 31, 2005."

8 NTT East claimed a market share for inter-city and intra-city calls of 64 percent and 71 percent respectively in 2003. See "Operating and Financial Review and Prospects," p. 9. 9 "Non-Consolidated Statements Of Income," Financial Statements for the F