‘Who invented the four modes of services supply?’

The idea evolved over almost three years among negotiators in the Uruguay Round talks in the 1980s, crystallised by the EU delegation

Photos representing the four modes

By Peter Ungphakorn
POSTED FEBRUARY 15, 2022 | UPDATED FEBRUARY 19, 2022

We now take for granted that services can be traded internationally in four ways known as the “four modes”, but once upon a time this was not so clear-cut.

The four “modes of supply” (or “modes of delivery”) are:

  • cross border supply — where a service provider in one country sells the service to a customer in another country without anyone moving, for example professional advice over the telephone or internet
  • consumption abroad — where the customer travels, for example tourism
  • foreign commercial presence — where the service provider sets up a subsidiary or branch in another country, for example a bank or insurance company
  • movement or presence of natural persons — where individuals travel to provide the service either as staff in the foreign branch or subsidiary or independently, for example maintenance engineers travelling to service aircraft abroad

This is now established right at the top of the World Trade Organization’s General Agreement on Trade in Services, under Article 1, “Scope and Definition”:

 2.   For the purposes of this Agreement, trade in services is defined as the supply of a service:

(a)  from the territory of one Member into the territory of any other Member;

(b)  in the territory of one Member to the service consumer of any other Member;

(c)  by a service supplier of one Member, through commercial presence in the territory of any other Member;

(d)  by a service supplier of one Member, through presence of natural persons of a Member in the territory of any other Member.

How they are now applied can be seen in the commitments that each WTO member has made in services, in fiendishly complicated documents known as “schedules” of commitments (explained in this technical note).

But it took some time to arrive at that point.

“Who invented the four modes?” asked economist Richard Baldwin on Twitter.


NOTE

This is not intended as a survey of literature, simply a description of how the discussion developed in the Uruguay Round, based on available documents from the time. It was the Uruguay Round that produced the formulation of the four modes as we know it, in Article 1 of the General Agreement on Trade in Services (GATS).

Clearly there were already discussions elsewhere as the 1987 observation from the UN Statistical Office shows. Several people have also pointed to “Identifying the Issues in Trade in Services” by Gary Sampson and Richard Snape from June 1985, before the Uruguay Round was launched. It seems likely that the article and the authors influenced the negotiations.

Their formulation is slightly different from the GATS outcome. They also have four modes. But one is movement of “factors of production”, which covers GATS modes 3 and 4 and more. Their fourth mode is “transactions [that] may occur with the movement of both factors of production and the receiver of the service”, in a third country. It is not one of the four GATS modes. In the Uruguay Round, “factors of production” features in 1988 and 1989, including the documents for and from the 1988 Montreal ministerial meeting.

Detailed replies came from international lawyer Amy Porges and Jonathan Scheele, Research Associate at European Studies Centre, St Antony’s College, Oxford University.

Both had considerable experience in the negotiations that created the WTO and its trade agreements and in developments since then. Porges was in the GATT Secretariat. Scheele served as one of the EU’s negotiators on services. A few others chipped in.

This account builds on those replies by looking through the official documents from the 1986–94 Uruguay Round negotiations under what was then the General Agreement on Tariffs and Trade (GATT).

The talks created the WTO to replace the GATT (as an international organisation), and expanded the multilateral trading system to include new areas of trade, including services.

The round was officially launched in Punta del Este, Uruguay, in September 1986. For developing the four modes, the key period recorded in official documents was from February 1987 to October 1989, two years and eight months.

The main relevant documents are from the services negotiating group in the round. They are in the archived GATT series of WTO documents, their code numbers beginning with MTN.GNS (MTN = multilateral trade negotiations; GNS = Group of Negotiations on Services).

Services were entirely new to multilateral trade rules and negotiators had to start from scratch. They had to grapple with fundamental questions ranging from the meaning of “trade in services”, to precisely which activities count as services.

Less than six months after the Uruguay Round was launched negotiators had started discussing the types of delivery.

Photo of person with laptop and phone to illustrate Mode 1: Service provided across the border
Mode 1: Service provided across the border, such as by phone or over the internet
February 23–25, 1987, meetings

The summary of these meetings (MTN.GNS/7) contains the earliest reference to the types of delivery.

In this extract, from page 7, one or more speakers refer to (1) cross-border supply, (2) (partial) presence of natural persons, (3) commercial presence. The note does not mention consumption abroad:

24. It was also said that the question of definition of trade in services would be central to the impact of an agreement on the world services economy. While it could be agreed that the concept should not be stretched to include all aspects of services transactions, a definition would nonetheless have to be sought which reflected the reality of the services economy and the objectives agreed in Part 71 of the Ministerial Declaration. An important issue was the identification of the various types of services transactions. For example some services, characterized as pure cross-frontier trade, could be wholly produced in one country and consumed by residents of another. Some services were partly produced in one country but also partly, during the temporary presence of an employee of the producer, in the consumer’s country. Some services were essentially produced within the central organisations of a producer in one country but contact with the consumer took place through some form of permanent commercial presence of the producer in the consumer’s country. The ability to produce a particular service could be sold to the resident of another country through a licensing or franchising agreement.

April 9, 1987, Singapore paper

In a paper on consumption versus trade (MTN.GNS/W/6), Singapore discussed these concepts in detail

April 8–10, 1987, meetings

The summary (MTN.GNS/8) of the meetings shows the types of delivery were discussed a lot under “definitional issues”. The first time consumption abroad was mentioned was in paragraph 12 (page 4).

12. One member referred to the view expressed by some other members according to which an important criterion was whether a service crossed national frontiers. However, in the case of tourism, for instance, the consumer but not the service was crossing national borders. Therefore if trade in services included tourism it could not be defined exclusively in terms of a movement of services across national borders. On the other hand, he agreed that if Ministers at Punta del Este had intended that the negotiations on services should also deal with all aspects of foreign investment in services they would have indicated this in their Declaration. However, in the field of services it was often difficult to clearly distinguish between trade and investment. It was therefore for the GNS to develop a suitable definition of trade in services. In his view it should opt for a definition which was sufficiently broad to enable the Group to fulfil the overall aims of the negotiations which were economic development and growth. As regards the classification of services proposed by one member he thought it would be useful to have an illustrative list of services sectors. With respect to personal services he said that since tourism was part of this sector it was not correct that personal services in general were non-tradeable services.

The way these modes were described in and around this meeting shows how much negotiators were still struggling with concepts.

One delegation proposed five types of delivery (one being purely domestic). Here, one of the complexities came from the concept of “residence” (from page 5).

14. One member suggested the following classification of transactions in services: first, there were services destined for consumption in the domestic market (e.g. real estate transactions, personal services, public services); these services, which accounted for more than ninety percent of total world production of services rarely crossed national borders. A second group of services consisted of services which were offered by residents of a country to non-residents without crossing national borders. Examples of this type of services were tourism and services provided to non-residents in harbours and airports. From the point of view of the payment for the service these transactions could be considered as exports. Thirdly, there were services offered by residents of a country to non-residents which crossed national borders (e.g. international transport, international consultancy services, communications etc.) A fourth group of services comprised services furnished internationally in the framework of contractual relations, such as representation agreements. A fifth group consisted of services which could be traded only on the basis of the establishment of a branch or subsidiary in the country where the service was to be consumed. This category of transactions caused a conceptual problem with respect to the definition of the term “residency”. If “residency” was defined on the basis of geographic location of an enterprise, sales of services by a foreign enterprise located in a country would be considered as part of the internal transactions in that country; if, on the other hand “residency” was defined with reference to the nationality of the enterprise such a transaction could be seen as an import of a service. This classification of services transactions showed that there were various ways in which services could be traded internationally. Each type of international service transaction raised particular problems with respect to the collection of reliable statistics. One example of such statistical problems concerned services transactions which took place between the parent and branch or subsidiary of a transnational corporation. Earnings of overseas branches and subsidiaries were reflected in the balance-of-payments under the heading “income from direct investment” but this item did not distinguish between investment in services and investment in goods producing sectors.

Photo of tourists, to illustrate Mode 2: people travel to consume
Mode 2: Consumption abroad, such as tourism
June 19, 1987, information from other organisations

The negotiating group also sought information from other international organisations. In a paper compiling their replies (MTN.GNS/W/9) the UN Statistical Office had this to say on the types of delivery (pages 38 and 39):

III. OBSERVATION PROBLEMS OF MEASURING EXTERNAL TRADE OF SERVICES

17. The traditional method used in the observation of exports and imports of goods, based on customs declarations in general, cannot be used for services. Services, at least most of them, do not cross the border in the same way as goods and escape customs declarations. The following main categories can be distinguished in the export and import of services:

(i) the buyer of the services crosses the border and purchases the services in a country in which he is a non-resident. For instance, tourists visit foreign countries and buy hotel-, restaurant-, hairdresser services;

(ii) the seller of the services crosses the border and sells his services in a country in which he is a non-resident. For instance, he or she goes to a foreign country and provides some consultancy services or carries out some repairs on machinery delivered earlier;

(iii) both the seller and buyer are outside the country of their residence when the service is provided. For instance, a PANAM flight carried Japanese passengers from New Delhi to Frankfurt;

(iv) both the seller and the buyer remain in their countries of residence, the service is provided by correspondence and there is some tangible or quasi-tangible end result of the service. For instance, a consultancy report or computer tape is prepared in the country of the seller and sent to the country of the buyer;

(v) as in (iv), but there is neither tangible nor quasi-tangible end result of the service provided. For instance, insurance service is provided to a non-resident, and there is nothing other than the payment (premiums, reimbursements) that crosses the border;

(vi) some goods or quasi-goods cross the border and the services are provided on them, or in connection with them. For instance, goods sent abroad for repair, transit transportation and transit telecommunication.

This shows that other organisations were also grappling with the concepts. Whether any of them ended up with four is not reflected in the Uruguay Round negotiating documents.

(Interestingly, points (iv) and (vi) of the UN Statistical Office’s comment link services with goods, or with a “tangible end result” such as a consultancy report in hard copy. In a slightly different way, EU economist Lucian Cernat has recently been advocating a “mode 5”, to take account of the value of services embedded in traded goods.)

Over the weeks and months discussions continued and the concepts developed. Sometimes “presence” was defined in terms of “factors of production”, which covered establishing a branch or subsidiary as well as the movement of personnel.

June 24, 1988, Secretariat paper

A GATT Secretariat paper on “definitions” (MTN.GNS/W/38/Rev.1) contained a long examination of two concepts:

  • cross-border sales of services where the service itself crosses the border
  • sales of services which include factor flows across the border and where the service itself does not necessarily cross the border.

Banks and logos to illustrate mode 3, commerical presence
Mode 3: Commercial presence, such as bank branches or subsidiaries
November 25, 1988, report to ministers

The Uruguay Round was supposed to end in 1990 (it missed the deadline and actually ended in 1994). A meeting of the round’s umbrella Trade Negotiations Committee was held at ministerial level in Montreal in December 1988 as a mid-term review.

The services negotiating group’s report to the Montreal meeting (MTN.GNS/21) referred to types of delivery in two separate parts, which may reflect alternative approaches from different delegations.

The section on “definition” (page 3) included three types of cross-border movement: services, consumers and factors of production (which would cover commercial presence and movement of people). The text was in square brackets [] indicating that it had not been agreed.

“Definition” is where the modes ended up more than five years later in the final agreement.

Definition

9. [Work should proceed on the basis that the multilateral framework [shall] [can] include trade in services involving cross-border movement of services, cross-border movement of consumers, and services for which the movement of factors of production is essential for the provision of services abroad. [Further] examination is needed on to what extent the framework would cover the movement of production factors such as capital or labour flows.]

[In particular, the multilateral framework will provide facility for foreign service providers to establish.]

[In particular, the multilateral framework will provide facility for labour, labour-intensive and professional service providers [from developing countries] to take residence [in developed countries] for the purpose of providing such services.]

[An approach to definition should take into account the following criteria:

(i) Cross-border movement of service and payment,
(ii) Specificity of purpose,
(iii) Discreteness of transaction,
(iv) Limited duration.]

Then, in the same report, came the first time that “mode” was used in this context (“mode of delivery”).

It came under “market access” (page 5), apparently the approach of the US and perhaps some others. It was also in square brackets, meaning it might also have been an alternative to the “definition” text.

[[Market Access] [Preferred mode of delivery]

[When market access is available to signatories, it will be on the basis that, consistent with other provisions of the agreement, foreign services may be supplied [, unless negotiated otherwise,] according to the preferred mode of delivery essential to market access, e.g. through movement of factors, through establishment or other forms of commercial presence and through the temporary movement of skilled or other qualified persons.]

December 9, 1988, ministerial decision

The Montreal ministerial decision, MTN.TNC/7(MIN) echoed the report to ministers by putting the delivery types under “definition”. But they were no longer under “market access”.

The decision was the official mid-term review of the Uruguay Round, setting the direction for the remainder of the negotiations. The services section was Part 2 (from page 40. MTN = multilateral trade negotiations; TNC = Trade Negotiations Committee; MIN = ministerial).

4. Work on definition should proceed on the basis that the multilateral framework may include trade in services involving cross-border movement of services, cross-border movement of consumers, and cross-border movement of factors of production where such movement is essential to suppliers. However, this should be examined further in the light of, inter alia, the following:

(a) Cross-border movement of service and payment.
(b) Specificity of purpose.
(c) Discreteness of transactions.
(d) Limited duration.

In 1989, discussions continued under the Montreal mandate as negotiators dug deeper into concepts and how to deal with them.

September 18–22, 1989, meetings

The negotiating group returned to paragraph 4 of the Montreal decision (“work on definition”). In lengthy discussion several delegations referred specifically to commercial presence and movement of people, instead of just factors of production. The meetings’ notes (MTN.GNS/25) are 110 pages!

October 17, 1989, US draft

A draft agreement from the US took a different approach (MTN.GNS/W/75). Chapter 3 on “Market Access” (pages 4–5) had separate articles on “Establishment”, “Cross-Border Provision of Services” and “Temporary Entry for Services Providers”. Article 17 on “general definitions” (pages 13–15) included movement of consumers, grouped with establishment and cross-border provision, and separately “provision of a service ‘by a person’”.

Photo of aircraft maintenance engineer to illustrate Mode 4: presence of person
Mode 4: Presence of personnel, such as aircraft maintenance engineers
October 19, 1989, EU paper

This was the first time the four modes appeared together clearly, although they were not called “modes”. It was in an EU paper on “definition of trade in services”, and constituted the entire proposed definition of trade in services (MTN.GNS/W/76 page 3).

Proposed approach to the Definition of Trade in Services

For the purposes of the multilateral framework, trade in services shall be considered to Include the provision of a service through one or more of the following types of transaction:

  • transactions involving cross-border supply of the service;
  • transactions involving cross-border movement of consumers;
  • transactions involving a commercial presence, whether temporary or permanent (including through the creation of wholly-owned subsidiaries, branches, representative offices, joint ventures, partnerships and franchising operations), providing that activities are limited to the specific purpose for which access was granted;
  • transactions involving the movement of personnel essential to the supply of the service (to be defined by sectoral annotation as appropriate) i.e. key personnel and other skilled personnel, and providing that such movement is limited to the specific purpose for which the access is granted, and is of limited duration or is a discrete transaction.

It is understood that liberalisation in any sector will be progressive and does not necessarily cover all forms of trade. The relationship between the scope of the framework and the specific liberalisation commitments/bindings of each signatory to the framework is set out in section C of the Community’s proposal “The Basis for the Progressive Liberalisation Process” (doc. MTN.GNS/W/66).

This approach ended up in Article 1 Paragraph 2 of the 1994 General Agreement on Trade in Services.

It’s pretty clear therefore that this formulation came from the EU. Jonathan Scheele says the EU was thinking about it before the paper was circulated, but it’s also clear that the ideas had already been percolating in the negotiating group for some time. He also described the services negotiating group as “very much a collegial operation”.


Based on this Twitter thread

Updates:
February 19, 2022 — adding the link to the technical note on services schedules of commitments
February 16–17, 2022 — correcting the reference to “modes of delivery” in the pre-Montreal report (it was under “market access” not “transparency”); adding the sidebar note on this article’s focus; adding the reference to “mode 5”

Image credits:
Mode 1 cross-border supply (service by phone and internet) | Mikhail Nilov, Pexels licence
Mode 2 consumption abroad (tourism) | Streetwindy, Pexels licence
Mode 3 commercial presence (banks) | Expect Best, Pexels licence
Mode 4 presence of personnel (aircraft maintenance engineers) | Delta News Hub, CC BY 2.0

Author: Peter Ungphakorn

I used to work at the WTO Secretariat (1996–2015), and am now an occasional freelance journalist, focusing mainly on international trade rules, agreements and institutions. (Previously, analysis for AgraEurope.) Trade β Blog is for trialling ideas on trade and any other subject, hence “β”. You can respond by using the contact form on the blog or tweeting @CoppetainPU

%d bloggers like this: