A draft manual, currently under consideration, for the collection of statistics and data on trade in services could in effect re-write the GATS framework, tilting the balance in favour of the delivery of services by 'commercial presence'.

by Chakravarthi Raghavan

Geneva, 4 Sep 2000 -- Unless developing countries and their trade establishments, negotiators at the WTO and trade and finance ministries back in their capitals, assert themselves now, they may find that the proposed methodology for collection of national and international data on services trade will in effect re-write the GATS framework and its definition of services, tilting the balance in favour of investment.

A UN-led effort to agree on a manual for collection of statistics and data on trade in services will not be ready for approval by the UN General Assembly before 2001, and even then the internationally-agreed manual will not fully reflect the definition of Trade in Services in all four modes of delivery set by the General Agreement on Trade in Services, but rather tilt in favour of delivery of services by ‘commercial presence’.

A draft manual under consideration, drawn up without active participation of statistical experts of developing countries, makes a reference to the four modes of supply under GATS - generally referred to as Mode 1 or cross-border supply, Mode 2 or consumption abroad, Mode 3 or commercial presence and Mode 4 or presence of natural persons.

“For the sake of clarity and statistical feasibility”, the July draft says, the manual does not “strictly rely” on the GATS legal provisions on modes of supply. To allow any particular transaction to a mode of supply, the manual complements the GATS definitions and proposes unambiguous criteria based on the “territorial location” of the transactor at the time the service is supplied as well as on the type of supplier, individual or business enterprise.

These criteria, it is claimed, yield the same results as the GATS definition “in most cases”. It does not spell out the cases where the same results are not achieved.

By following the BPM5 methodology, in trade in services between residents and non-residents, the manual excludes financial intermediation services indirectly measured (FISIM). The argument is that this is done for practical reasons, since “some financial intermediaries provide services for which they do not charge explicitly by paying or charging different rates of interest to borrowers and lenders.”

The draft notes that, in principle, almost any mode of supply is conceivable for any type of service. But to facilitate analysis and compilation of service transactions by modes of supply, the manual has suggested a “simplified approach”.

Identification of modes of supply, considered relatively marginal in a given service category, is not recommended.

“Thus if the major part of a given service category corresponds to services supplied under one mode, this mode could be fully allocated in that category. For e.g. freight transport (which may actually involve multimodal transport, within and outside a country) would be allocated to cross-border supply. “A given service category would normally correspond only to one or two modes of supply”.

The manual acknowledges that there are cases where specific Balance of Payment (BOP) service transactions involve several modes of supply - e.g. when an architect designs a construction project and delivers it over a telecommunication network and has to make visits to the country of the consumer at the implementation stage. If estimates cannot be provided for the subdivision of transaction value by modes, “the transaction be allocated to the most important mode in terms of resources and time associated with it.”

In this approximation, construction services (involving commercial presence) for a short-term construction project, will be wholly assigned to Mode 3. And the BOP component of construction service will include transactions in Mode 4.

The manual suggests that all transport, communications, insurance and financial services, together with payment of royalties and license fees, be allocated to Mode 1.

All services recorded in BOP as travel are to be deemed consumption abroad or Mode 2.

This, in fact will give a misleading picture. Some estimations show that in fact under tourism, hardly 10-15% of the benefits (mostly wages of low-level labour) go to the benefit of the country receiving tourists, while most of the benefits are retained or go to the country or firms where they originate. Moreover, the allocation of financial services to Mode 1 seems arbitrary since the subject-matter of negotiations under the GATS has concerned principally Mode 3.

The collection of data even on this inadequate basis would not begin until after 2001, and it may probably take 3-4 years before proper data can be assembled, and developing countries, and their trade and finance ministries and central banks, would be able to make a judgement on the effects of their market-access commitments, and the benefits they have received for their own exports.

This means that in terms of the work at the World Trade Organization in the new round of negotiations on trade in services, they would have undertaken new commitments, without a proper assessment of the effects of their commitments in the past, or an estimation of costs and benefits of their new commitments.

At the minimum, it would seem that they must use their leverage in the services negotiations now to ensure that the manual for services data collection to be agreed upon, be not based either on the IMF Balance-of-Payments statistics manuals or even the Extended BOP manual, which would reflect only international transactions in terms of transactions between ‘residents’ and ‘non-residents’ as defined by the IMF.

But even if they do, and succeed in ensuring that the agreed methodology of collection of statistics fully conform to the definition laid out in the GATS agreement, and the four modes of delivery, they will still not have any real data from international sources to enable them properly to assess the effects of the market access commitments under the GATS so far, and thus cannot estimate what would be the effect of any new commitments or the balance of concessions, in the new round of service negotiations that have been launched.

Some financial experts, like Mr. Andrew Cornford, a senior economist at UNCTAD specializing on financial market issues, say that developing countries may need to develop their own expertise (with some assistance from outside) and develop methods of assessment, using national data.  For this purpose, they might have recourse to “real options” theory.

Options theory in financial economics is concerned with valuing options (the right to sell or buy stocks or other financial instruments at a price specified in the contract, the “strike” price), and what is now the key formula for this purpose was developed by Fisher Black, Myron Scholes and Robert Merton, leading eventually to the award to the last two of the Nobel Prize in economics. Originally designed to value stock options, the theory was also applied to other financial instruments and commodities and, following the work of Michael Brennan and Eduardo Schwartz, to a wider range of economic decisions in which an entity disposed of, and needed to value, an economic right such as a gold mining concession or a patent, which it was none the less under no obligation to exploit.

According to Conford, this theory could in principle be used to estimate the value of offers or commitments in trade negotiations.Key variables in the theory are the value of the asset to which the option applies and its variance, the strike or exercise price, the time until the expiration of the option, and a riskless interest rate. The problems of extending the theory to offers or commitments are difficult but probably not insurmountable.

In such a case, for the value of the asset in “real options theory”, one would require some proxy such as the revenue from the business covered by a sectoral commitment or offer in the country making the offer or commitment; one would also require an estimate of its variance; and for the strike price one would need the cost of entering and engaging in that business.(In the case of a gold mining concession discussed in the original article of Brennan and Schwartz, for instance, the proxy for the strike price was the cost of opening the mine and the present value of the cost of producing gold.)

Cornford emphasized that the involvement of Scholes and Merton in the failure of the hedge fund, LTCM, has no bearing on the utility of “real options theory”, which is an attempt to address problems for valuation and decision making presented by contingent opportunities (of which offers and commitments no less than gold mining concessions and patents are examples). Moreover, he does not underestimate the practical difficulties of extending the application of the theory to the new area of valuing offers and commitments in trade negotiations. However, this approach would have the advantage of requiring sectoral data at a national level which in principle should exist and, he opines, more often than not does exist, and not data on international transactions and income which currently are not available and seem unlikely to become so in the foreseeable future.

Over the longer term, this approach will be no substitute for gathering proper and comparable and published data, nationally and internationally on cross-border service transactions. But the real option approach could be used as a starting point, and perhaps negotiators could build into their offers the right to withdraw or modify (without having to pay any compensation to trade partners), when proper data are available, or if the data collection process makes no progress say, in 4-5 years more.

And until some approach such as that based on the 'real options' theory is developed, Third World countries may do well not to undertake any new commitments or even agree to old modalities (of requests and offers, or binding at the WTO/GATS their autonomous liberalization) in the new round of negotiations.

Even as it is, developing countries lost out in the Uruguay Round and its outcomes, in that there has been no fulfilment of the overall promises when the Round was launched that in return for extending the trade system’s remit in new areas like intellectual property and services, they would get better market access in goods (agriculture and textiles), that in the service sectors and modes of delivery, there would be a balance between capital and labour, and there would be market opportunities to export in sectors and sub-sectors where developing countries have advantage.

As Mr. Bhagirath Lal Das, former Indian representative to GATT, puts it, a fraud was perpetrated on developing countries in terms of the promised enhanced market access in the areas of textiles and clothing, and agriculture.

And the major benefits of the WTO/GATS and the further services negotiations launched at the WTO (in the financial and basic telecommunications services sectors) have benefited the major industrialized countries, particularly the US and Europe.

But if developing countries, who did not pay much attention to these problems during the Uruguay Round, even now undertake new commitments in services, on the basis of economic theories (that the industrialized countries don’t themselves practice) about the relation between trade liberalization and increased ‘welfare’ or about global efficiency rewarding the most efficient producers through “free trade”, they may be in for more problems in the future.

But this time, at least, neither negotiators, nor their governments back home, would be able to tell their parliaments and public that they did so out of ignorance, and there was no alternative.

The issue of lack of services data, and the inadequacies of data collection using the IMF Balance-of-Payments statistics, was raised by developing countries as early as 1986 (in the run-up to the Uruguay Round) and again in April 1987, and several times thereafter. During the Uruguay Round itself, UNCTAD and UN statistical divisions took up this issue.

The statisticians came to the negotiators, and asked them to provide a definition of ‘trade in services,’ to enable the statisticians to agree upon a methodology for collection of national data, and their collation at international level. They also underlined that all this would require enhanced international cooperation and allocation of sufficient resources.

But before they could proceed further, the issue of agreed methodology for collection of data was hijacked by a small group, the so-called Voorburg group (after the place in Netherlands where they first met), mostly from OECD, IMF, World Bank and Eurostat (the statistical wing of the European Commission).

At a planning meeting of the UN statistical office in New York at that time, when the UN raised the issue of Third World participation, they were told that the problem should not be complicated by bringing in the Third World representatives.

The issue came up again in 1990 and in 1994, at the GATT and at UNCTAD.

But nothing very much happened.

Subsequently, while the UN/UNCTAD restructuring (before the UNCTAD-IX meeting at Midrand) resulted in downgrading of the statistical work at UNCTAD, its work in the area was also focused (by the inter-governmental body) on the problems of data relating to movement of natural persons.

At present, the issue of establishing a manual for collection of data, is being handled by a task force of the OECD, IMF, World Bank, Eurostat, the UN, and UNCTAD - with the latter represented by a relatively junior professional.

UNCTAD itself is now trying to repair the situation by upgrading work on statistics and recruiting a competent senior statistician. But given the UN system of recruitments and promotions (with various gender, internal promotion, and regional representation and other preferences taking priority), it is not clear how long it will be before a person is appointed, and how long it would take for the person to catch up on this work (and how willing he will be to stand up against those pursuing the interests of the developed countries and their corporations).

The draft manual on ‘Statistics of International Trade in Services’, prepared jointly by the European Commission, IMF, the OECD, UN, UNCTAD and the WTO, was considered by an expert group at its meeting in New York on 10-12 July.

In the light of the comments, the draft is to be revised, and a new draft is expected to be sent by 15 September for comments, and a final draft prepared by 15 October.

This is first to be considered by an OECD expert group on 11 December, to be followed by a meeting of the task force on 14 December (both to be held at the OECD).

The manual has to go before a meeting of the UN Statistical Commission, probably early in 2001, and then has to go before the UN General Assembly itself. The Statistical Commission group is said to be planning an ‘informal preparatory planning’ meeting in Washington in November.

Developing countries and their statisticians and experts won’t be involved until the issue comes before the Statistical Commission, and they may well be told that any changes they seek now would put back the data collection process, that in any event the collection of data reflecting fully all the four modes of delivery is difficult and complicated, and one has to adopt rough approximations.

In the draft that was before the July expert group meeting, in a general introduction and overview, the draft recognizes the diverse needs of services data, including the need to support the GATS negotiations and their implementation, and says: “To meet these diverse needs, and in recognition of the role of affiliated enterprises abroad in the delivery of services and the increasing tendency for commercial agreements to cover this method of supply, the term ‘international trade in services’ is construed broadly in the Manual. It covers trade in the conventional sense of transactions between residents and non-residents and the value of services delivered through locally established enterprises.... designated (in the manual) as Foreign Affiliates Trade in Services.”

While featuring some important steps forward, the manual does so “by building upon, rather than by suggesting modifications to, internationally agreed standards (emphasis added) for compilation.”

The standards used are those in the fifth edition of the IMF Balance of Payments Manual (BPM5), containing recommendations for the definition, valuation, classification and recording of resident/non-resident trade in services, and the System of National Accounts 1993 (1993 SNA), whose concepts underpin the proposed draft Manual’s recommendations concerning data on services delivered through foreign affiliates.

A box in the manual explains that while the BPM5 concept of services, in principle, is essentially same as that of the SNA, “for practical measurement reasons”, the international trade between residents and non-residents includes some trade in goods, such as those bought by travellers and similar goods purchased by embassies. On the other hand, payments for international trade in goods may under certain circumstances indistinguishably include service charges, such as insurance, maintenance contracts, transport charges, royalty payments, packaging and software.”

This really means that outgoing service payments on these heads will not be reflected in the data.

“For services delivered through subsidiaries and branches abroad, referred in the manual as commercial presence, methodological antecedents are not fully developed,” says the July draft. But drawing on the work conducted by the OECD and Eurostat, as well as experience of a number of countries collecting such data, “the Manual reflects the emerging international consensus that statistics on such services should be developed for firms in which a foreign investor has a majority interest, and should be classified primarily on an activity basis (i.e. by industry of the producer rather than by type of service produced.)”.

But this would be contrary to the classification of services used for negotiations under the GATS.

The GATS definition of ‘commercial supply’, and the commitments scheduled, do not have such a requirement.

Thus instead of being guided by what the negotiators have agreed and set in the WTO/GATS rules, the small group of statisticians of the North are determining how services data should be developed.

Based on industry groupings drawn from the International Standard Industry Classification (ISIC) of All Economic Activities, and Revision 3 (ISIC) used in reporting statistics to international organizations, the ISIC categories for Foreign Affiliates (IFIA) allow activities of services enterprises to be viewed in the context of activities of all enterprises.

“Although detail by product for foreign owned firms would be desirable to enable comparability between GATS data and trade between residents and non-residents, compilation on product basis will remain a longer term goal owing to current limitations on data collection,” says the draft by the task force.

As for the mode of supply of natural persons, the definition and concepts used in the GATS require some information outside the BPM5 and FATS domains or that pertain to transactions that the BPM5 records in components other than services. And because these domains are not viewed as subject to modification, the data on supply by natural persons is to be addressed in a separate annex outside the main manual.

The level of detail (for collection of data) is presented as “a compromise between the need of information of trade negotiators, analysts and policy makers and difficulties of data collection” However, trade experts note, that the ‘commercial presence’ definition in the GATS range all the way from a mere representative office for a foreign service provider abroad, through various categories of joint ownership and participation in capital of locally incorporated firms and branches, to wholly-owned foreign firms.

Neither the IMF’s BOP nor the EBOP manuals and methodologies, based on the resident-non-resident concepts, fully reflect this situation.

The draft says: “For services delivered through subsidiaries and branches abroad... methodological antecedents are not fully developed."

Thus, more than a decade since this issue was raised in the context of the Uruguay Round negotiations, developing countries may find themselves agreeing to commitments without a clear assessment of their value or their effects on the balance of payments.

In the area of international trade in goods, there are rough ways of ascertaining or estimating the value of concessions offered and exchanged - through an estimation of the recent value of goods exported, the tariff rate in the importing country and the tariff cut offered, and an estimation of the price elasticity of those goods.

Trade theorists and economists can and do find fault with these bilateral and reciprocal mercantalist views of the concessions, but it is an available measure of sorts. But, as explained above, this is lacking in the area of services.

There are no reliable data on international trade in services that could provide a basis at present for countries to make offers and exchange concessions. The IMF’s Balance of Payments data, or its Extended Balances of Payments data, do not provide such a basis - since they are based on a definition of international transactions as one between a resident and a non-resident.

Under the GATS, trade in services has been defined as involving supply in four modes: supply of services from one country into another (cross-border); supply of service in a country to service the consumer of another (consumption abroad), by a service supplier of one country through commercial presence in another country (commercial presence), and by a service supplier of one country member through presence of natural persons in the territory of another (presence of natural persons).

The efforts to agree upon and put in place a method for collection on a commonly accepted basis of data on trade in services, nationally in all countries, that could then by collated into figures for international trade, has made little progress so far.  And developing countries appear to be particularly at a disadvantage as the major industrialized countries, promoting the interests of their transnational corporations as the dominant service suppliers, in combination with some trade officials, appear to be proceeding on the basis of getting developing countries to commit themselves in GATS, and locking them into often disadvantageous positions, before they can fully assess and be aware of the implications.

The developing countries could have the option of not moving further by way of making offers or liberalising until a satisfactory system of collection of data reflecting all four modes of delivery is put in place and data covering at least five years is available to provide a valid basis for judgement by their capitals.

Alternatively, and as a stop-gap, they could try to use the ‘real options’ theory to make some rough estimates of the future value of any offers or concessions they make or get.-SUNS4735

The above article first appeared in the South-North Development Monitor (SUNS) of which Chakravarthi Raghavan is the Chief Editor.

© 2000, SUNS - All rights reserved. May not be reproduced, reprinted or posted to any system or service without specific permission from SUNS. This limitation includes incorporation into a database, distribution via Usenet News, bulletin board systems, mailing lists, print media or broadcast. For information about reproduction or multi-user subscriptions please e-mail < >