(Reprinted from HKCER Letters, Vol. 7, March 1991)
Making Ends Meet -- The Other Side of International Telecommunications: A Comment on Dr. Mueller's International Telecommunications in Hong Kong
Dr. Mueller's Case
Dr. Milton Mueller's short book has been a big hit locally, and deservedly so. It is written in a vigorous and uncompromising style, making the points unambiguously and with force. Its central message is undoubtedly correct: Greater choice through competition should be the aim of telecom policy. And so is its secondary thesis, the one that has attracted most attention, that international services are very profitable, and there are ways to pass more of those profits onto users through tariff reductions. So what's the problem?
The problem is that the diagnosis and the cure do not fit together as easily as Milton implies. Let us start with the diagnosis.
International telephone charges, what are called "collection charges," consist in essence of two components. First, the costs really incurred in transmitting a call, which are comprised of an allocation of fixed costs, like labor costs, depreciation, and HKTI's (Hong Kong Telecommunications International Ltd.) share of international facilities such as its contribution to Intelsat for its satellite segment, and variable costs which are mainly switching costs. The secondcost comes from the profit margin, on which HKTI pays royalties in addition to taxes. In 1987 royalties came to HK$156 million.
Now HKTI has protested that there is a third cost element, namely the accounting rate payment which all receiving telephone administrations pay out to sending administrations. If a particular traffic stream is favorable to Hong Kong, that is, more call-minutes are incoming than outgoing, HKTI receives a net income from overseas. This is true of the U.S.A.-Hong Kong stream, but not of any other. In all other cases, incoming call-minutes are less than outgoing, and HKTI pays a net outflow of dollars. In my view, as an economist, HKTI is wrong to call these outflows a cost, but from HKTI's accounting viewpoint, the fact that they think of this outflow as a cost is understandable. (In fact, HKTI is not alone among telecom administrations in adopting this procedure. For example, British Telecom does so.)
I agree with Milton on this point. By a cost an economist means a cost which cannot be avoided under any circumstances if the service is to be delivered. In the case of the accounting rate levy, the accounting rate itself is just a payments convention between administrations. It could, in theory, be abolished tomorrow, and most users wish it would be because it appears like a tax on their telephone bills. To HKTI it is a genuine cash outflow, but an outflow from their revenues. It is a form of revenue-sharing in which overseas administrations receive part of the collection charge levied on Hong Kong users.
The reason the accounting rate system is so pernicious is that it is long out-of-date. In the early days of its operation it did largely reflect the genuine costs of making a telephone call, as Milton points out. But as costs the world over declined, mostly due to new technologies, accounting rates stuck. A far better system would be for each telephone administration to keep its own collective charges, but this raises the question, what incentive does it leave for an overseas telephone administration to receive and transmit messages? An interim solution is to lower existing accounting rates.
Who benefits from the present system of high accounting rates? (To be accurate, we should talk of the accounting rate share which is the proportional split between administrations, usually 50-50, and what that share is 50-50 of, which is the rate.) In terms of international revenue flows, the answer is countries which enjoy a traffic stream balanced in their favor. Generally, in poor countries, collection charges are high, which encourages users of all kinds to arrange their calls as incoming. Poor countries earn a lot of scarce foreign currency in this way. For a deficit administration, like HKTI, the consideration must be that the deficit on any one stream does not rise above the revenue from local collection charges for that stream. Overall, there is clearly no danger of that in Hong Kong, but on individual traffic streams, it could and sometimes does happen. (I believe it happens on the Taiwanese stream, for example).
It is clearly nonsense to hope that poor countries will be persuaded easily to give up this vital source of currency earning. But between highly developed and world-integrated economies there are grounds for optimism. If between them accounting rate levels were substantially lowered towards real costs, then collection charges could also be lowered, and if, as is generally expected, international telecom traffic is price-elastic, then total traffic volumes will grow even faster than at present. Smaller margins on higher turnover will compensate telephone administrations for lower tariffs.
Milton would disagree with none of this, I am sure, but how to achieve this desired result? And what local consequences would flow from it? Milton has only one proposal here: competition. He would have the Hong Kong government revoke the HKTI international voice licence (which is not due to expire until 2006). Let us suppose that were done. A second network enters the field, called Twoco, and begins cutting international telephone charges. We all rejoice, of course. But who does Twoco link up with overseas to receive the calls? If it is a PTT monopoly, then the accounting rate levy will still be there. If it is a duopolist in the U.K. or triopolist in Japan or the U.S.A, can Twoco negotiate a lower rate? With AT&T, MCA or Sprint the answer may well be yes, but with BT and MCL? Or KDD and IDC or ITJ?
Regulators in both Britain and Japan have followed America in having rules (known as "uniform settlements" or "parallel accounting") against "whipsawing," which happens when an overseas administration tries to beat down the accounting rate in negotiations with one national carrier, using the threat that it will send its traffic via the competing national carrier unless agreement is reached. No overseas carrier has any incentive to do this unless the traffic stream revenue is unfavorable to them, which means favorable to the receiving country.
The significance of this point was specifically raised last November by Britain's Department of Trade & Industry, in its otherwise pro-competition telecom policy statement, Competition and Choice. I quote:
A potential consequence [of whipsawing] is that UK carriers would have to charge higher prices to their customers than would otherwise have been the case in order to pay their share of the accounting rate. (par. 10.17)
This illustrates one complexity of the problem. And this concern came not from a leftist government but from a Thatcherite regime. Milton Mueller's implied view that a second international carrier (how many should there be?) will resolve the problem of accounting rates I find unconvincing. By shaving its margin of profit, a second carrier will no doubt reduce international tariffs, at least for the business sector, with the predictable consequence that more traffic will originate from Hong Kong, and the net outflow of payments from Hong Kong will rise. HKTI will certainly lose profit share, and the gainers, besides Twoco, will be mainly corporate businesses and overseas telecom administrations.
For me, the most interesting part of Milton's book was his analysis of how easily small-business users and residential customers would also benefit from reduced international call charges. Taking what he says is a reasonable assumption that local telephone rental would have to rise by 20 percent (in 1989 residential rentals were $576 per annum) to offset a reduced average international tariff (he assumes the IDD $9.5 per minute) of 25 percent, he estimates the average telephone user would only need to make an additional 48 minutes of international telephone calls a year to break even at the new tariff levels. First, this assumes that all IDD call charges would fall, not just those offered to the business sector. It is not clear why this should be so if we assume that a second network in Hong Kong would only choose to serve the corporate business centers of Central and Tsim Sha Tsui. HKTI would only need to respond in those market sectors.
Second, from this it follows that the intervention of a regulator might still be necessary even after Twoco has entered the arena. In Britain, Oftel recently decided to intervene to price cap BT's international tariffs despite the presence of a Twoco. In fact, a regulator could intervene with HKTI whether Twoco was allowed in or not. The result Milton is positing is easily accomplished if international tariffs become part of the regulator's concern. Under the present Scheme of Control, revenues from franchised services are allowed to reach a rate of return on shareholders' funds of 16 percent, so a reduction in IDD revenues would require an increase in local rentals. Taking figures for 1987, I estimate the loss of revenue to Telco from such an IDD tariff reduction at around $240 million, which would imply an average local rental increase of $120 per year, or 23 percent on local residential rentals (which were then $516 per year).
A third point worth bearing in mind is that it is likely that local rentals will have to increase fairly substantially for other reasons, even if the present Scheme of Control is replaced by a pricecap. So local sensitivity and resistance to rate hikes may grow.
If the Scheme of Control is replaced by a price-capping formula, which seems the most likely, then tariff rebalancing is bound to follow. To hold local rentals within bounds, some implicit cross-subsidization from other service revenues may continue. International services are the most likely source, but all value-added services, international and local, are possible sources. Competition will cream-skim the most lucrative value-added markets, and either the regulator will have to work out some burden-sharing formula through which new entrants make a contribution to the costs of the social obligations placed upon the telephone company, or the telephone company itself will be allowed to charge discriminatory prices which favor the competitive sectors and disadvantage the residential and possibly the small-business sectors.
In this latter case, new entrants will complain about predatory pricing policies, and the regulator will become an arbitrator. In all these scenarios, competition poses as many problems as it solves. This is not an argument against competition, or the threat of it, but an appeal to recognize the realities of a complex mesh of contending interests, what we may call the political economy of telecommunications.
Private and Social Benefits
In telecommunications studies, the traditional emphasis has been on issues of allocative efficiency (do relative prices reflect relative costs?) and technical efficiency (can low prices still be profitable?). There is now a need to become more aware of the growing strategic importance to an economy like Hong Kong's of the dynamic efficiencies of telecom. Dynamic efficiencies refer to questions of facilities and services innovation, quality and range as well as price. Telecom is no longer just POTS (the plain old telephone service), it has been transformed by information technology and now offers information services which are directly productive of wealth, not least as trade in services. Competition as a policy issue in this context really arises with respect to whether barriers to entry, and to exit, inhibit innovation. Most, if not all, of the tendencies in the market which Milton outlines in the last section of his book, are likely to happen. Companies with private networks will want to use them to communicate directly with customers without using the public switched network. Companies with spare leased capacity will want to resell it. Companies will want to hook up directly to international gateways. Carriers with alternative transmission facilities, such as mobile, CT2s, PCNs, microwave links, will offer competing services. Certainly, many companies will find it cost effective to bypass the public network, but Milton overstates the case. He does so because he wrongly, in my view, characterizes each single circuit on the network as conceptually a different service:
A network that adds new subscribers or extends itself to new locations is not producing more 'units' of the same service, it is producing a different service. (p.30)
If this were true, it would imply that a call to one's mother, or to the police emergency service, which comes from a telephone box is a different service from exactly the same message being sent from home. Milton is understating, if not entirely denying, the public good aspect of the telephone network. The larger the network, the greater the benefit it offers to all individual subscribers. This is the reason why private bypass will never mean that someone leaves the public network. If it did mean such a thing, then the private benefits derived by the person or company quitting the network would have a social cost. By understating, or denying, the public good aspect of telecommunications, Milton is overstating the capacity of competition to achieve an optimal market solution. It is unlikely that telecom, with or without competition, can ever be left just to the market if society perceives telecom as having a role to play in a community's social and economic goals.
John Ure is an economist/telecommunications specialist and Research Associate at the Centre of Asian Studies at the Hong Kong University. He has been providing consultancy services to Cathay Counsellors Group Inc., which counts among its clients Hong Kong Telecom. However, the company cannot be considered in any way responsible for, and may not be in agreement with, points of view expressed in this article.