UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT

POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES

STUDY SERIES No. 2

E-COMMERCE, WTO AND DEVELOPING COUNTRIES

by

Arvind Panagariya

Professor of EconomicsCo-director, Centre for International Economics

Department of EconomicsUniversity of MarylandCollege Park MD 20742

UNITED NATIONS

New York and Geneva, 2000

NOTE

The views expressed in this study are those of the author and do not necessarily reflect the viewsof the United Nations.

The designations employed and the presentation of the material do not imply the expression of anyopinion whatsoever on the part of the United Nations Secretariat concerning the legal status of anycountry, territory, city or area, or of its authorities, or concerning the delimitation of its frontiers orboundaries.

Material in this publication may be freely quoted or reprinted, but acknowledgement is requested,together with a reference to the document number. A copy of the publication containing the quotationor reprint should be sent to the UNCTAD secretariat:

ChiefTrade Analysis Branch

Division on International Trade in Goods and Services, and CommoditiesUnited Nations Conference on Trade and Development

Palais des NationsCH – 1211 Geneva

UNCTAD/ITCD/TAB/3

UNITED NATIONS PUBLICATION

Sales No. E.00-II-D-23

ISBN 92-1-112494-8

ISSN 1607-8291

Copyright 8 United Nations 2000All rights reserved

iii

ABSTRACT

In this paper, I discuss the policy issues e-commerce raises for the World TradeOrganization (WTO) and developing countries. I advocate three policy prescriptions.First, it will be most appropriate to classify e-commerce as trade in services withGATS discipline applied to it. Developing countries should ensure that e-commerceis not classified as goods trade with a permanent zero custom duty pact. Such an out-come would liberalize all e-commerce by default, undermining their bargainingpower.

Second, at present there is some disagreement about whether international Inter-net transactions should be classified as cross-border trade or consumption abroad. Inmaking their commitments in the UR and post-UR negotiations in services, countriespresumably viewed these transactions as cross-border trade. Therefore, Internettransactions would be best classified as cross-border trade.

Finally, developing countries such as India that have the capacity to exportskilled services through Internet should aggressively negotiate market access with de-veloped countries in the future WTO negotiations. This involves negotiations on twofronts. One, they should seek liberalization by developed countries in sectors inwhich they have a comparative advantage. And two, they should seek recognition oftheir education, qualifications, requirements met, or licenses or certificates granted inthe markets of other countries.

Electronic commerce offers unprecedented opportunities to both developing anddeveloped countries. In the short run, the gains are likely to be concentrated in deve l-oped countries but, in the long run, developing countries have more to benefit. This isbecause, in the short run, developing countries lack the infrastructure necessary totake full advantage of Internet. But in the long run, they can leap frog, skipping someof the stages in the development of information technology through which developedcountries have had to pass.

Key Words: e-commerce, Internet, WTO, developing countries.

iv

v

ACKNOWLEDGEMENTS

I am deeply indebted to Susan Teltscher for discussions on virtually all aspects ofthe paper as it progressed. The paper benefited from incisive comments by Bijit Boraand Aaditya Mattoo and from discussions with Jolita Butkeviciene and Erich Supper.The inspiration to work on e- commerce came from John Cuddy who kindly invitedme to be a Visiting Scholar in the Division on International Trade in Goods andServices, and Commodities (DITC), UNCTAD.

vi

vii

CONTENTS

ABSTRACT.................................................................................................................iii

I. INTRODUCTION ..............................................................................................1

II. WHICH MULTILATERAL DISCIPLINE: GATT, GATS OR BOTH? .....3

III. MODE 1 OR MODE 2? .....................................................................................8

IV. ACCESS TO E-COMMERCE........................................................................10

A. Access to internet services ...........................................................................10

1. Availability of infrastructure, hardware and software ............................102. Access to communications networks ........................................................10

B. Access to electronically traded services.......................................................12

V. INTELLECTUAL PROPERTY RIGHTS.....................................................14

VI. E-COMMERCE AND DEVELOPING COUNTRIES .................................15

A. The gains from internet to developing countries..........................................15

B. Policies for the expansion of e-commerce ...................................................19

VII. CONCLUSIONS...............................................................................................23

1

I. INTRODUCTION

There are currently six differentmediums of electronic commerce (e-commerce): telephone, fax, television,electronic payment and money transfersystems, Electronic Data Interchange(EDI) and Internet.1 It is fair to saythat even though phone, fax and televi-sion remain the most widely usedelectronic mediums to promote or con-duct commerce, much of the currentexcitement, confusion and debate on e-commerce is the result of the rapid as-cendancy of Internet. Internet hasmade international transmission ofservices possible in ways and on ascale via traditional modes such as fax,phone and television. It is being usedtoday to buy abroad many back-officeservices such as electronic publishing,website design and management, cus-tomer call centres, medical recordsmanagement, hotel reservations, creditcard authorizations, remote secretarialservices, mailing list management,technical on-line support, indexing and

1 Created by the trucking industry in theUnited States in the early 1970s, EDI entailsthe exchange of documents and informationbetween the computers of two businesseswithout human intervention. Stores such asWalMart use the technology to link their sup-pliers directly into their stock databases.Through the link, suppliers are automaticallynotified and authorized to send shipmentswhen the shelves are bare. According to theEconomist (May 10, 1997), 95 per cent of theFortune 1,000 companies use EDI.

abstracting services, research and tech-nical writing, and technical transcrip-tion. Internet has also become a me-dium for electronic transmission ofmany products, traditionally traded inthe form of goods. For instance,books, CDs, movies and computerprogrammes can now be transmittedinternationally in digital form.

From the viewpoint of multilateralrules of international trade as well asnational economic policy, this mediumgives rise to issues somewhat differentfrom those faced with respect to othermediums. For instance, WTO mem-bers must decide whether the GATT orGATS discipline should be applied tointernational trade via Internet. To theextent that some of the trade via thismedium has a counterpart that is tradedphysically, as is true of books, com-puter programmes, music and movies,one may apply the GATT discipline.But to the extent that such counterpartsdo not exist, as is the case with theback office services mentioned above,it will make more sense to apply theGATS discipline. From the viewpointof national economic policies, espe-cially in developing countries, the po-tential for development, offered by thismedium, increases the urgency to de-velop the telecommunications industryand create the financial infrastructurethat facilitates electronic transactions(for example, credit cards).

2

In the present paper, I discussthese and other aspects of e-commercefrom the viewpoint of developingcountries.2 In Sections 2-5, I offer ananalytic discussion of multilateral ruleslikely to be applicable to Internetcommerce. Special attention is paid toissues of taxation and access to e-

2 The reader may find it useful to acquiresome background information on various elec-tronic mediums from the more comprehensivestudy, Electronic commerce and the role of theWTO, Special Studies 2 (Geneva: World TradeOrganization), 1998. Additionally, theEconomist has published two detailed surveyson e-commerce in issues dated May 19, 1997and June 26, 1999.

commerce. In Section 6, I focus on theimplications of e-commerce for deve l-oping countries and discuss possiblepolicy measures the countries maywish to take in order to maximize thebenefits from it. The paper is con-cluded in section 7.

3

II. WHICH MULTILATERAL DISCIPLINE: GATT, GATS ORBOTH?3

3 In writing this section, I have benefited greatly by e-mail exchanges with Aaditya Mattoo and ac-cess to his ongoing research with Ludger Schuknecht for a forthcoming paper entitled "Trade Policiesfor Electronic Commerce.”

The degree to which countries canregulate international trade via theInternet, what taxes they can imposeon it, and in what way they can dis-criminate in favour of the domesticsuppliers of similar items will dependon the WTO discipline the membercountries decide to apply to it. TheWTO report mentioned in footnote two[WTO (1998) henceforth] raises thepossibility that, in principle, the “dig-its” traded on Internet could be viewedas goods, services or even somethingelse. Which of these characterizationsis chosen determines whether this tradeis subject to the rules laid down in theGeneral Agreement on Tariffs andTrade (GATT), General Agreement onTrade in Services (GATS), a combina-tion of these two or an entirely newagreement.

It may be noted at the outset thatthere is no ambiguity at present re-garding the status of the goods orderedand paid for on Internet but deliveredphysically in the conventional manner.Except for the order and paymentthemselves, these transactions aretreated as goods trade and the GATTdiscipline applies to them. The ambi-guity arises only when the goods aredelivered on Internet.

On the face of it, any deliveriesmade by Internet would seem to re-

semble services. Nevertheless, as al-ready noted in the introduction, thereare products delivered by Internet thathave counterparts in physical, mer-chandise trade. The obvious examplesare books, videos, music CDs andcomputer software. When imported inphysical form, these products aretreated as goods with the GATT disci-pline applied to them. But can they betreated as services when delivered byInternet? Or, in conformity with theirphysical counterparts, should they betreated as goods?

One extreme possibility is to cha r-acterize all transmissions on Internet asgoods with GATT discipline applied tothem. Such a characterization accom-panied by a ban on custom duties onthe transmissions, currently in place,would amount to the WTO memberscommitting themselves to completefree trade in all transactions routed byInternet. This is because nationaltreatment and MFN status are generalobligations under GATT. By accept-ing the GATT discipline, under na-tional treatment, the member countrieswould give up their right to discrimi-nate against Internet imports as far asdomestic taxes are concerned. In ad-dition, the ban on customs duty wouldbind their tariffs on Internet imports atzero. However, at present, no one isconsidering such a proposal. The

4

member countries made their commit-ments in the UR and post-UR negotia-tions on services based on the assump-tion that most of those transactionswere services rather than goods.

At the opposite extreme, we couldabandon both GATT and GATS anddevelop an entirely new discipline forInternet trade. Once again, virtually noone is advocating this position. For asearch for a new discipline for e-commerce makes little sense. Internetservices, which include Internet serviceproviders and phone lines on whichtransmissions flow, are already subjectto GATS and the Agreement on BasicTelecommunications. All electronictransmissions that flow on Internet, onthe other hand, have counterparts ineither goods trade or services trade.As such, the rules necessary to regulatethat trade can be found in GATT orGATS.

Thus, the real choice is betweenapplying GATS to all Internet trade, orGATT to that trade for which physicalcounterparts also exist, and GATS toall other e-trade. In my judgement, onbalance, it makes more sense to defineall electronic transmissions as services.At one level, it may be argued that atthe time Internet transmissions crossthe border between two countries, theydo not have a physically traded coun-terpart. The eventual transformation ofthe transmission into a good such as abook or CD does not negate the factthat at the border the transmission didnot have a physically traded counter-part. Indeed, in many cases, thetransmission may not be turned intothe physically traded counterpart at all.For example, the recipient may con-tinue to store it in the digital form withbooks read on the screen and musicplayed directly on the computer.

But this is not the primary reasonwhy I lean in favour of treating allInternet trade as service trade. The keyadvantage of adopting the across-the-board definition is that it is clean andminimizes possible disputes that mayarise from countries wishing to havecertain transmissions classified as in-tangible goods and others as services.Under a mixed definition, in any tradedispute involving Internet trade, panelswill have to first decide whether theobject of dispute is a good or a serviceto determine whether the rules ofGATT or GATS are to be applied inevaluating the dispute. The adoptionof the across-the-board definitionautomatically resolves this issue.

The across-the-board definition,nevertheless, raises some efficiencyissues that must be addressed. Thus,consider first the issue of tariffs, whichare applicable to products imported inphysical form but not when transmittedelectronically. As long as the cost ofelectronic transmission is lower thanthat of physical delivery, the presenceof tariffs on the latter poses no prob-lem. Effectively, the electronic trans-mission offers the product to the coun-try at a price lower than that availablethrough physical delivery. Thischange is equivalent to an improve-ment in the country’s terms of tradeand, leaving aside some general-equilibrium considerations, improveswelfare unambiguously.

But for many countries, especiallydeveloping ones, this is an unlikelyscenario. In these countries, most con-sumers do not have computers orInternet access. A likely scenario,therefore, is one in which a handful ofindependent entrepreneurs will receivethe product by Internet, convert it intophysical form such as CDs and sell thelatter to consumers. But this activity

5

may itself be costly, using up real re-sources.

A possible outcome of the pro-posed regime in many developingcountries can be represented stylisti-cally, therefore, with the help of Figure1. In the figure, DD gives the demand

for a specific compact disc (CD) andGG its supply when imported in phys i-cal form, as a good. It is assumed thatthe country is small so that the supplyis perfectly elastic. In the absence ofInternet transmission, the quantity pur-chased is given by Q0 and tariff reve-nue by ABGGt.

Suppose we next introduce Inter-net transmission. Assume, as is truecurrently, that if music is transmittedelectronically, no tariff is paid. Com-petitive entrepreneurs import musicelectronically, convert it from digitizedform into CDs and sell them to con-sumers. The marginal cost of conver-sion and distribution is positive andrising, leading to the supply curve EE.It is then immediate that quantity OQewill now be imported by the electronicmedium with QeQ0 continuing to comein physical form. The tariff revenue

collected previously on the quantityOQe disappears. Of the lost revenue,area marked 1 goes to cover the highercosts of supply by Internet and is adeadweight loss. The remainder of thelost revenue becomes a transfer to ex-porters.

This is the standard story from thesmuggling literature that arises whenthere are two sources of supply and themore expensive source is not subject toa tariff but the less expensive source is.It should, of course, be clear that if the

Figure 1

Price

QuantityO

D

D

G G

Gt Gt

E

E

1

Qe Q0

A

B C

6

cost of Internet transmission were lowsuch that the Internet supply curvecrossed the demand curve below GG,this problem would not arise. Internetsupplies will eliminate physical ship-ments and the price will be below GG,benefiting the consumers by more thanthe lost tariff revenue.

This analysis shows that subject-ing like products, delivered by differ-ent means, to different disciplines canpotentially result in harmful efficiencyeffects. This is not an inevitable out-come, however. There are at least twosolutions to the problem. First, thecountry could choose to eliminate thetariffs on physical deliveries, thuseliminating the efficiency loss such asthat represented by area 1 in Figure 1.Indeed, this will lead to a net effi-ciency gain of triangle ABC. Second,if the tariff on physical supplies cannotbe eliminated because of fiscal consid-erations, the country could choose toimpose a higher VAT or excise tax onmusic CDs supplied by Internet by anamount equal to the tariff on physicaldeliveries. As long as the country hasnot already committed itself to givingnational treatment to imported musicservices, this option is available withinGATS.4

It is useful at this point to returnbriefly to the temporary ban on cus-toms duties on all electronic transmis-sions mentioned earlier. While thisban would be meaningful if all e-

4 The absence of trade taxes on services isan important and entirely neglected problem.If tariffs are imposed to raise revenue, effi-ciency dictates that services are brought intothe tariff net as well. Yet, this issue has re-ceived no attention in the academic or policyliterature presumably because academics stilllike to think of services as non-traded andpolicy analysts do not want to scare away for-eign investors by taxing the services suppliedby foreign sources at higher rates.

commerce were classified as goodstrade, its continued existence and thecurrent United States proposals tomake it permanent are puzzling. Atpresent, the only feasible method ofcharging a customs duty on electroni-cally supplied foreign services is tosubject them to a higher domestic taxrelative to the identical, domesticallysupplied services.5 As long as a coun-try has not committed itself to givingnational treatment to the foreign serv-ice in question in its national schedule,it is free to impose a higher domestictax on electronically supplied servicesfrom abroad. The existing ban oncustoms duty and the United Statesproposal to make the ban permanent,do not and cannot forbid countriesfrom subjecting an imported service toa higher VAT or excise tax than anequivalent domestically supplied serv-ice. The discriminatory treatment isforbidden only if the Member commitsto giving the imported service nationalstatus in its national schedule. But inthat case, the current ban on customsduty and the United States proposal tomake it permanent have no additionalimpact. In either case, the ban ismeaningless and entirely vacuous.6

A second difference betweenGATT and GATS discipline from theviewpoint of efficiency, is that theformer does not allow quotas while the

5 Even this option is available in the caseof business-to-business transactions only.When a foreign business sells a product elec-tronically directly to domestic consumers, it isnot clear how the transaction can be subject toany domestic taxes.

6 One possible explanation is that in theSeattle Round negotiations, the United Statesmay still be intending to get Internet tradeclassified as goods trade. And if by then thecountries have already committed to a perma-nent ban on customs duty, Internet trade willautomatically be freed of all border restric-tions.

7

latter does. In the particular example Ihave discussed above, in principle, ifWTO members decide to apply GATSdiscipline to services traded electroni-cally, a country will have the option tolimit the number of CDs that could betransmitted by Internet. It is not im-mediately clear how this restriction canbe enforced. But assuming that it

could be done, trade will be diverted toshipments in physical form, which maybe an inferior mode of delivery. Atpresent, such a quota is not enforce-able. If it does become enforceable,the outcome can be inferior to that ob-tainable under the GATT discipline.This will be a cost of the clean defini-tion I have advocated.

8

III. MODE 1 OR MODE 2?

The General Agreement on Tradein Services classifies services accord-ing to the mode of delivery. It distin-guishes four modes: cross-border sup-ply (mode1), consumption abroad(mode 2), commercial presence (mode3), and the movement of natural per-sons (mode 4). Assuming the GATSdiscipline is applied to electronic trade,for transactions that do not take placeeither through commercial presence orthe movement of natural persons, themember countries will still need to de-cide whether they are to be treated ascross-border trade (mode 1) or con-sumption abroad (mode 2).7 There areno clear-cut objective criteria that canbe brought to bear on this classifica-tion. Therefore, it is likely to be nego-tiated as a part of the next round of ne-gotiations. The choice of classificationhas two principal implications.

First, the classification will deter-mine the liberalizing impact of thecommitments made in the UR andpost-UR GATS negotiations on serv-ices. In these negotiations, countrieshave already made commitments basedon the modes of supply of services.Therefore, it matters whether elec-tronic trade is treated as being supplyby mode 1 or mode 2. For example, ifa country gave full market access un-

7 Though the discussions on e-commerceare often focused on cross-border method ofdelivery, it can and does take place throughcommercial presence (mode 3) as well as themovement of natural persons (mode 4). Forexample, when a foreign bank offers electronicbanking services to the residents of a country,the transaction is classified under mode 3.Likewise, when computer programmers moveto another country and offer their serviceselectronically there, such e-commerce will beclassified under mode 4.

der mode 2 for a particular financialservice that is traded electronically, thecommitment would have no liberaliz-ing impact if electronic commerce isclassified as supply under mode 1rather than 2. Thus, the liberalizingimpact of previous commitments willdepend on the mode supply underwhich electronic commerce is classi-fied. It is my impression that countriesundertook more obligations for liber-alization under mode 2 than undermode 1. Accordingly, the liberalizingimpact of the commitments will begreater if electronic commerce is cla s-sified under mode 2. Developedcountries, which are net exporters ofelectronic services, stand to gaingreater market access if these servicesare classified as being supplied undermode 2.

Second, the classification deter-mines the country of jurisdiction forpurposes of regulation and dispute set-tlement. For supply under mode 1, thetransaction is deemed to have takenplace in the country where the buyerresides. Therefore, it is the regulatoryregime of the importing country thatapplies to the transaction. In contrast,for supply under mode 2, the relevantregulatory regime is that of the countrywhere the supplier resides. If countriesfeel that they want to protect their buy-ers’ interests, they are likely to opt formode 1. Thus, there is some tension inthe choice of classification dependingon the objective. The market accessobjective pulls towards mode 2 whilethe consumer protection objective pullstowards mode 1.

To the extent that in making theirliberalization commitments in the URand post-UR negotiations, countries

9

viewed the electronic transactions be-tween providers and recipients in dif-ferent countries as cross-border trans-actions, it makes sense to treat them as

such. Otherwise, actual liberalizationis likely to result in being at variancewith what the countries intended.

10

IV. ACCESS TO E-COMMERCE

Access to e-commerce, which inthe WTO parlance often means accessto e-exports, has two components thatmust be distinguished sharply: accessto Internet services and access to serv-ices that can be traded electronically.The former deals with access to Inter-net infrastructure while the latter re-lates to specific commitments in elec-tronically tradable services (forexample, commitments in financialservices under modes 1 and 2). Ingoods trade, we can liken these com-ponents, respectively, to access totransportation networks (includingports, ships, roads, railways and airtransport) and access to specific goodsmarkets through a lowering of tradebarriers such as tariffs and quotas. Forlower trade barriers to result in moreimports, access to transportation ne t-works is necessary. Similarly, for spe-cific commitments in various servicessectors under modes 1 and 2 to resultin increased flow of imports, access toInternet facilities is essential.

A. Access to internet services

The access to Internet infrastruc-ture depends on two factors: (i) avail-ability of communications networks,hardware and software and (ii) accessto the existing communications net-works. Let us consider briefly each ofthese factors.

1. Availability of infrastructure,hardware and software

At the basic level, access to Inter-net by the residents of a country de-

pends on the level of development ofthe telecommunications sector and theavailability of hardware and software.In the remote villages of many deve l-oping countries, even the basic tele-communications services may not ex-ist. To bring Internet and, hence, e-commerce to these villages, one willneed first to bring telecommunicationsservices there. But even when tele-communications services exist, addi-tional hardware that links up the ind i-vidual user to Internet must be put inplace. Finally, one needs to ensure ac-cess to equipment such as computers,modems and software. Generallyspeaking, an open trade regime withrespect to information technologyequipment is likely to facilitate accessto this equipment. This is perhaps thereason why some countries chose tosign the Information TechnologyAgreement (ITA), which requires thesignatories to open up trade in a largenumber of information-technologyproducts.

2. Access to communications net-works

There are three principal WTOprovisions that govern access to com-munications networks: GATS ArticleVIII on monopolies and exclusiveservice suppliers, GATS Annex onTelecommunications, and the Refer-ence Paper on regulatory principles inthe Agreement on Basic Telecommu-nications. In addition, specific com-mitments on national treatment andmarket access made by countries in thebasic telecommunications sector have

11

implications for access to Internet.8

GATS Article VIII and the Annex ap-ply to all WTO members uniformly.The Reference Paper applies to ap-proximately 60 countries that incorpo-rated it into their specific commitmentsin the agreement on basic telecommu-nications services. A total of 69 coun-tries made specific commitments in thebasic telecommunications sector. Ofthese, ten countries made specificcommitments with respect to Internetaccess providers.

Article VIII, which applies to allservices, is designed to deal with mo-nopoly suppliers who can potentiallyfrustrate a Member’s MFN and spe-cific market access commitments. Forinstance, suppose telephone lines in aMember country are owned by a singleentity and the Member has made mar-ket access commitments to othercountries in the provision of Internetservices. Article VIII requires this en-tity not to limit access to phone lines toservice suppliers from other Membersor discriminate among them. It alsorequires this entity to ensure that thecommitments made by the Member inother service sectors are not frustrated.

Article VIII is limited in its appli-cation to cases in which a monopolistsupplies the service in question.GATS negotiators recognized, how-ever, that basic telecommunicationsservices are central to the smooth flowof trade in a large number of otherservices. Therefore they introducedfurther provisions in the Annex onTelecommunications to widen accessrights in the use of public telecommu-

8 For completeness, mention may also bemade of GATS Article IX on business prac-tices, which provides for consultation and in-formation exchange between affected Mem-bers when suppliers resort to anti-competitivepractices.

nications transport networks and serv-ices (PTTNS).9

The Annex requires each Membergovernment to ensure that suppliers ofother Members are given reasonableand nondiscriminatory access to anduse of PTTNS for the supply of aservice included in the Member'sschedule. The term "nondiscrimina-tory" is defined here to include bothnational treatment and MFN. The An-nex, thus goes beyond Article VIII intwo respects. First, for a service listedin the Member's schedule, it gives for-eign suppliers nondiscriminatory ac-cess to PTTNS even though the Mem-ber has not committed to nationaltreatment in that service.10 Second, theaccess provision applies to PTTNS ir-respective of whether these servicesand networks are supplied by a mo-nopolist or competitive firms.

The concern that telecommunica-tions markets would be dominated bylarge operators, capable of frustratingmarket access commitments, remainedcentral during basic telecommunica-tions negotiations. This led the par-ticipants to lay down a set of regula-tory principles in a Reference paper,

9 As defined in the Annex, a public tele-communications transport 'service' is any tele-communication transport service, offered tothe public, involving the real-time transmissionof customer-supplied information without anyend-to-end change in its form or content. Pub-lic telecommunications transport 'network'refers to public telecommunications infra-structure permitting telecommunications be-tween and among network termination points.

10 This means that if a country lists internetservice supplies in its national schedules evenwithout committing to national treatment, fo r-eign suppliers are to be given nondiscrimina-tory access to PTTNS. Discrimination againstforeign suppliers is still possible in other areas(for example, taxation) as long as the countryhas not committed to national treatment ininternet.

12

aimed at reigning in the behavior of themajor suppliers of telecommunicationsservices. Some 60 participants incor-porated this Reference Paper into theircommitment schedules.

The regulatory principles in theReference Paper oblige major suppliersto provide interconnection on nondis-criminatory terms. They are to providealso services in sufficiently unbundledform that those seeking interconnec-tions do not have to pay for unneces-sary components and facilities. TheReference Paper also lists rules gov-erning anti-competitive cross-subsidization, the misuse of informa-tion, licensing criteria and transpar-ency.

Finally, Internet access also de-pends on the degree of liberalization inbasic telecommunications undertakenby Members. 69 countries signed theAgreement on Basic Telecommunica-tions in February 1997. Counting theEuropean Communities as one, thisproduced 55 schedules. Many of thenegotiated undertakings represent apre-commitment to liberalize in thefuture.

A key area of liberalization fromthe viewpoint of Internet access is thatof Internet Service Providers (ISP). Inmany countries, telecommunicationsservices are supplied by a public mo-nopoly, which often also becomes themonopoly provider of Internet access.In countries, which have liberalizedtheir communications regimes, com-peting ISPs exist and offer differentbundles of Internet services. In futurenegotiations, it will be worthwhile toincorporate ISP as an explicit sectorinto national schedules of commitment.This may induce further liberalizationin many countries in this key area.There is no compelling argumentagainst permitting multiple ISPs or

foreign entry even in countries withmonopoly provision of other telecom-munication services.

B. Access to electronically tradedservices

In addition to the Internet accessservices just discussed, Internet offersthe opportunity for trade in two addi-tional areas. First, many services out-side of the telecommunications sector,such as those in the banking, insuranceand computer programming sectorscan be delivered electronically. Sec-ond, Internet can be the vehicle for theprovision of distribution services withgoods and services purchased throughInternet but delivered by other means.For transactions in the first category,GATS discipline applies fully. Incontrast, transactions in the secondcategory are similar to those by tele-phone or mail order. When deliveredphysically, goods are subject to theusual GATT discipline including cus-toms duties.

While national treatment and mar-ket access commitments in nationalschedules do matter in that they re-strain the importing country’s ability todiscriminate in its tax policies in fa-vour of domestic suppliers or amongvarious foreign suppliers, in the case ofInternet trade, they play a less crucialrole. To the extent that governmentsdo not have effective control over whatis traded on Internet, especially whentransactions are from business to con-sumers, the value of these commit-ments is limited.

Instead, the bulk of the expansionof e-commerce will depend on coun-tries granting recognition to the educa-tion or experience obtained, require-ments met, or licenses or certificatesgranted in another country. Article VII

13

of GATS allows for such recognitioneven on a discriminatory basis, in thesense that it allows Members to extendsuch recognition on a selective basis.For instance, the United States maygive recognition to accountancy de-grees from Europe but not India. Thiscould signal potential buyers that it ishazardous to buy accountancy servicesin India even though the latter may becapable of supplying them competi-

tively. In this regard, Article VII givessome flexibility to excluded countries,which developing countries should ex-ploit as much as they can. In particu-lar, if a Member gives recognition tothe standards prevailing in anotherMember in a specific area, and a de-veloping country’s standards in thesame area happen to be at par, underArticle VII provisions, it should begranted similar recognition.

14

V. INTELLECTUAL PROPERTY RIGHTS

The Trade Related IntellectualProperty Rights (TRIPs) Agreementapplies as much to transmissions onInternet as through other mediums.Copyright, trademark and geographicalindications must be respected in Inter-net transmissions the same way as inother mediums. In December 1996,two new treaties came into existenceunder the auspices of the World Intel-lectual Property Organization (WIPO),which deal specifically with Internettransmissions. These are WIPO Copy-right Treaty and WIPO Performancesand Phonograms Treaty. These trea-ties are to enter into force three monthsafter 30 countries have deposited theinstruments of ratification or accessionwith the Director General of WIPO.

The new WIPO treaties furtherstrengthen the rights of authors, per-formers and phonogram producers.The treaties recognize the role thattechnological measures used by rightsholders have in facilitating effectiveprotection. A variety of technologiesthat help control access or limit copy-ing of work transmitted via electronicmeans already exist and are being con-tinuously developed. The signatoriesto the treaties must provide adequatelegal protection and effective legalremedies against the circumvention ofthese effective technological measuresused by authors, performers and pro-ducers of phonograms.

Technologies also exist for incor-porating into digital copies of works

and other material, digital envelopesand watermarks that identify the work,its author and any other right holders,the terms and conditions of use of thework, and any other information. Thetreaties require signatories to provideadequate and effective remediesagainst any person, who alters or re-moves such information or distributescopies of protected material knowingthat such information has been re-moved without authority.

At present, these WIPO treatieshave not come into force. Howeverthey can be eventually expected to bebrought into WTO and incorporatedinto TRIPs. This may pose a problemfor developing countries given theircapacity to enforce disputes. In manydeveloping countries, courts have al-ready been stretched well beyond theircapacity and it is unlikely that they willbe able to deliver developed-countrystandards in the area of enforcement.As may happen with the existing en-forcement provisions in TRIPs, meet-ing the standards of developed coun-tries will give foreign rights holders afavoured treatment relative to domesticrights holders who will likely be sub-ject to the domestic pace of disputeresolution. Developing countries willneed to take into account these consid-erations, and possible threats of thedenial of Internet access by developedcountries, in making their decisionsregarding these treaties as and whenproposals are made to incorporate theminto WTO.

15

VI. E-COMMERCE AND DEVELOPING COUNTRIES

It is perhaps not an exaggeration tosay that, from the viewpoint of com-merce, Internet is the most importantinventions of the last two decades.This medium of "transportation" hasopened markets that were previouslyclosed, speeded up transactions as noother medium has done in the past, andmade the delivery of some productsalmost instantaneous.

In this section, I discuss the issuesmore directly relevant to developingcountries. I begin with an analytic dis-cussion of the ways in which Internetgenerates benefits for the countries andinteracts with other modes of deliveryof services, especially the movementof natural persons. I then considerpolicy actions that developing coun-tries may consider taking to enhancethe benefits from e-commerce.

A. The gains from internet to de-veloping countries

While virtually all countries standto gain from the opportunities offeredby Internet, according to one view, de-veloping countries stand to gain morefrom it than developed countries. Theargument is that these countries are farbehind developed countries in terms ofinformation-technology infrastructure.Given the cost savings offered byInternet technology and the relativeease with which it can be provided,they can now skip several stages oftechnological development throughwhich developed countries had to go.Stated differently, developing coun-tries are much farther inside the currenttechnological frontier and, therefore,have larger potential benefits frommoving to it.

In the long run, this is a defensiblestatement. But it must be acknowl-edged that the benefits of e-commerceare distributed unevenly not onlyacross countries--both between andamong developing and developedcountries--but also over time. Giventhat three fourths of the current e-commerce is concentrated within theUnited States, perhaps this singlecountry has benefited most from it. Incontrast, for many poor countries inAfrica, the telecommunications infra-structure is so poorly developed that itwill take a long time before they areable to benefit significantly from e-commerce.

The benefits from e-commerce toa particular developing country, bothdomestically and internationally, de-pend on the volume of demand for, andsupply of, goods and services that canbe potentially traded on Internet. De-spite all the excitement surroundingInternet, it is likely that for many de-veloping countries the demand andsupply factors do not promise largegains, at least in the foreseeable future.Due to a lack of electronic means ofpayment such as credit cards, pay-ments will still have to be made byconventional means. This factor aloneis likely to limit considerably the scopeof domestic electronic transactions.Moreover, the domestic demand forservices that are electronically deliv-ered is likely to be limited. Due to lowcosts of internal movement of naturalpersons, even businesses, which haveheavy needs for customized software,are likely to rely on the physical pres-ence of personnel. In these countries,even if Internet were widely available,e-commerce, as distinct from email

16

and other communications, will not bea big success immediately.

In assessing the potential benefitsfrom international e-commerce to acountry, analysts often focus only onthe goods and services that it can ex-port. This is an incorrect approach,however, since benefits can arise froma reduction in the cost of imports asmuch as from an increase in the pricereceived for exports. Even if a countrydoes not export any services, it canbenefit from imports of services, pay-ing for them in terms of goods.Cheaper availability of medical, engi-neering and architectural services,long-distance learning and reducedcosts of transactions can confer bene-fits even if the country does not imme-diately export the services tradedthrough Internet.

To the extent that Internet effec-tively opens markets that were previ-ously closed, it is tempting to think ofit as another form of trade liberaliza-tion. But, in fact, it is much more: itamounts to a technical improvementthat lowers costs of transactions and, assuch, generates far larger benefits thanthe triangular efficiency gains fromtrade liberalization. Indeed, the de-cline in costs increases potential bene-fits from trade liberalization in manyservices sectors.

Among developing countries, thecountries best situated to benefit frome-commerce through export expansionare those with a substantial pool ofskilled labour, capable of working onor near the frontier of computer tech-nology. The case of India, which isalready benefiting from e-exports in abig way, best illustrates this point.

I had long held the view that Indiahad greatly over-invested in highereducation. At one extreme, the mosttalented individuals left the country in

search of better opportunities abroadand, at the other extreme, the countryhad with a large pool of educatedworkers whom the economy could notabsorb. Even today, the lowest-levelclerical jobs attract large number ofapplications from graduates and post-graduates.

The advent of computer techno l-ogy in general and Internet in particu-lar threatens to prove my view to beincorrect, however. The migration ofsome of the country's most talentedindividuals to developed countriesnotwithstanding, the country has theworld's second largest pool of Englishspeaking scientific manpower. Eachyear, Indian universities graduate asmany as 115,000 engineers. This pool,Internet and the opening to direct fo r-eign investment by India have com-bined to yield annual exports of asmuch as US$ 4 billion. 11

Because the international move-ment of natural persons is subject tosevere restraints, the value of marginalproduct of skilled labour in developedcountries is far higher than in deve l-oping countries. Though numericalestimates are not available, the poten-tial gains from the increased mobilityof natural persons are astronomical.Developing countries in general, andIndia in particular, have long sought arelaxation of restrictions in developedcountries on the movement of naturalpersons. But they have not achieved anotable success in this effort.

By making the sales of skilled la-bour abroad, possible without actually

11 This information was provided by De-wang Mehta in his presentation at the WTOconference “Potential for Electronic Co m-merce for Businesses in Developing Coun-tries” on February 19, 1999 and summarized inthe WTO document WT/COMTD/18.

17

moving natural persons physically,Internet has at last brought developed-country demand for skilled labour todeveloping countries. This has re-sulted in a large capital gain on the in-vestment India has made in highereducation during the last four decades.Thus, what had seemed to be a poorallocation of resources for decades, expost, promises to turn into an excellentinvestment.12

Figure 2 gives an analytic repre-sentation of the benefits from theopening of the market for skilled la-bour through Internet. For simplicity,divide the world into two countries andcall them the United States and India.Use an asterisk to distinguish the vari-ables of the United States from thoseof India. Let M*M* represent the po-tential excess demand for skilled la-bour in the United States and EE theexcess supply of it in India. In view of

12 The simultaneous liberalization of directforeign investment has also helped this proc-ess. The presence of foreign firms in India hasplayed an important role in linking the demandfor various services in their source countrieswith the supply in the host country (i.e. India).

the fact that the United States is verylarge in economic terms, M*M* isshown to be relatively elastic.

In the absence of Internet and themovement of natural persons, skilledwages in the United States and Indiasettle at WA

* and WA, respectively.The introduction of Internet allows"trade" in skilled labour between theUnited States and India provided theUnited States has opened up its im-ports of some services through modes1 and 2. To the extent that Internet isan imperfect substitute for the move-ment of natural persons, and trade inservices under modes 1 and 2 is notentirely free, we will not expect theequilibrium to move to the fully inte-grated equilibrium, I. Instead, trade islikely to be limited up to, say, Q1, gen-erating gains from trade equal to thearea between M*M* and EE overquantity OQ1.

18

The important question is howthese gains are going to be dividedbetween the United States and India.The answer to this question depends onwhere the wage settles. When naturalpersons are allowed to move, the an-swer is clear. The wage is determinedon the demand curve, M*M*. This isbecause the American firms mustcompete for the limited number ofworkers who have been granted entryvisas. It is also the case becauseUnited States laws do not permit localfirms to hire foreign workers at a wagelower than that paid to United Statescitizens to ensure that firms do not optfor the former because they can em-ploy them cheaper.

The outcome is likely to be differ-ent when Internet is the medium of ex-ports of skilled labour. Now the wagewill be closer to the export-supply

curve, EE. This is because the wagemust be determined within the Indianmarket based on how much can be ex-ported. The more liberalization inservices the United States undertakesunder modes 1 and 2, the greater thedemand for Indian skilled labour andthe higher the wage. Thus, benefits toIndia depend directly on the extent ofliberalization undertaken by the UnitedStates in services that can be poten-tially exported by India on Internet.

This analysis is, of course, highlystylized. Cross-border trade will notsubstitute for the movement of naturalpersons in all cases. Often confident i-ality or security considerations requireconsultants to move to the site wherethe service has to be provided. Themost striking recent example relates tothe Y2K contracts. In other circum-stances, the movement of natural per-

Figure 2

Services of Skilled Labor

Skilled Wage

M*

M* E

EWA

WA*

Q1O

WT

WT* I

19

sons may even be complementary toexports via Internet. For instance, in-stallation and maintenance of softwaremay require the physical presence ofthe supplier. Finally, natural personsmay also be employed in sectors thatremain largely non-traded. This isclearly true, for instance, in the case ofmedical and health services.

We may also ask whether trade onInternet might substitute for direct for-eign investment. Sometimes it is sug-gested that if delivery by modes 1 and2 becomes a substitute for delivery bymode 3, Internet will become a substi-tute for direct foreign investment. Al-though examples of modes 1 and 2 de-liveries substituting for mode 3deliveries are not pervasive, this doesnot rule out the possibility that Internetmay have an adverse impact on directforeign investment. Substitution be-tween modes impacts only the sectoralcomposition of direct foreign invest-ment, not its aggregate level. Instead,the aggregate level will depend onwhether Internet raises the return oncapital more in the source countries orthe host countries. If the former, as islikely to be the case, at least in theshort run, more capital will choose tostay in the source countries. This isclearly an empirically testable hy-pothesis and is worth studying further.Internet has expanded sufficiently al-ready in developed countries for itsimpact on investment abroad to be de-tectable in the data.

B. Policies for the expansion of e-commerce

Development of e-commerceshould not be treated as a goal in itself.Some countries are better positionedthan others to achieve a rapid expan-sion of e-commerce for the sameamount of resource invested. Since

resources have alternative uses, onemust compare the rate of return in e-commerce to those in other activitiesbefore committing resources to thissector. This consideration remainsvalid even if investment decisions aremade by private agents, but the poli-cies chosen by the government havesignificant effects on those decisions.For instance, policies facilitating thedevelopment of e-exports are likely toyield higher returns in a country likeIndia, which has a significant pool ofskills to export, than in a countrylacking such skills.

For developing countries that findthe expansion of e-commerce a desir-able instrument for achieving its socialand development goals, action must betaken at three levels. First, the hard-ware and software necessary to de-velop electronically sellable servicesshould be available at reasonableprices. Second, the basic infrastructurenecessary for the smooth functioningof Internet must be in place. Here “in-frastructure” is defined broadly andincludes facilities to conduct financialtransactions on the Internet. Finallyand most importantly, developingcountries must negotiate access to de-veloped country markets in sectors towhich they can export services byelectronic medium. Let me take eachof these areas in turn.

Countries can ensure the access tohardware and software by liberalizingthe imports of the relevant products.This, in turn, can be accomplished byeither signing the Information Tech-nology Agreement or liberalizing theimports of the relevant products selec-tively, outside of that agreement. Notethat this recommendation is made tak-ing as given, the desirability of the ex-pansion of e-commerce in the firstplace. We must bear in mind thatwhen there are high trade barriers on

20

other products, as is likely in many de-veloping countries, this liberalizationitself may misallocate resources andconsumer expenditure. In such cir-cumstances, the benefits from the ex-pansion of e-commerce must outweighthe costs of the misallocation.

It is presumably in the area of in-frastructure development that deve l-oping countries need to do most to as-sist in the development of e-commerce.Without adequate telecommunicationssystems and the availability of inex-pensive telephone services, Internetand e-commerce cannot flourish. Atpresent, the telecommunications net-work in many developing countries israther poorly developed. A large ma-jority of individuals do not have accessto telephones,13 and those who do mustpay very high rates for telephone calls.Unlike in the United States, local tele-phone calls are metered and charged atfairly high rates so that even if theInternet access is cheap, the expense oflocal telephone calls, necessary to con-nect to the internet access provider, canraise the overall cost of Internet use.

There is also the issue of powersupply. In India, for instance, publiclysupplied power has been so unreliablethat many software firms in Banaglorehad to resort to their own generators toensure a continuous flow of power.Frequent and long interruptions inpower flows can have a devastatingeffect on the transmission of data.

13 In China and India, there are 2.3 and 1.1telephones per one hundred inhabitants. Thiscompares with 59.5 telephones per hundredinhabitants in the United States. Among de-veloping countries, only Hong Kong and Sin-gapore have telephone availability that is com-parable to that in developed countries. SeeTable 2, p. 7, WTO (1998), op. cit.

At present, in the large majority ofdeveloping countries, Internet access isalso expensive and unreliable. Oftentelecommunications services are sup-plied by a public monopoly, which alsobecomes the monopoly provider ofInternet access. Unable to expandservice sufficiently, under public pres-sure, it makes many more connectionsthan the capacity of the system. Theresult is a failure of many customers toaccess the service for which they havepaid.

The solution to this problem is tosimply allow private Internet serviceproviders into the market. As long asthese access providers can be obligedto give inter-connections to one an-other through proper regulation, thereare no benefits to having a monopolysupplier of the access service. This isclearly an area in which the privatemarket can function efficiently.

The prevalence of a legal frame-work, centred on paper-based contractsand handwritten signatures can alsoimpede the growth of e-commerce.The United Nations Commission onInternational Trade Law (UNCITRAL)had drawn attention to this issue asearly as 1985 and called upon govern-ments to consider the possibility ofpermitting, where appropriate, the useof electronic means of authentication.Subsequently, UNCITRAL developeda Model law on Electronic Commerce,which was approved by the UnitedNations General Assembly in Decem-ber 1996. The Model law lays outwhat constitutes the equivalent of awritten document, signature and origi-nal in the electronic environment. Italso sets forth rules governing the ad-missibility and evidential weight ofelectronic messages, the retention ofdata messages, the formation and va-lidity of contracts, and attribution.Many countries have either adopted the

21

Model law or introduced legislationrelated to electronic facilitation issues.The countries that have not yet intro-duced legislation along these lines arelikely to need to do so.14

Finally, assuming the provision ofreliable Internet service at reasonablerates domestically can be ensured, ad-ditional policy measures are requiredto facilitate e-commerce. In many de-veloping countries, electronic means ofpayment, including credit cards, arevirtually non-existent. This means thateven when products can be ordered orservices delivered by Internet, paymentmust be made by conventional means.This slows down the completion oftransactions considerably, reducingpotential benefits.

In the case of foreign purchases,this problem becomes even moreacute. Many developing countries donot have current-account convertibilityso that ordering goods on Internet fromabroad is not a practical option exceptperhaps in the case of large firms,which may have ready access to for-eign exchange. Even in countries suchas India, which have current-accountconvertibility but not capital-accountconvertibility, individuals do not haveready access to foreign exchange.Thus, as far as imports of goods andservices are concerned, the Internetoption is likely to remain limited tolarger firms. The solution is not en-tirely clear since the issue of givingaccess to foreign exchange to individu-als has serious implications for theability to control capital outflows, es-pecially in times of crisis. Even if ac-cess is provided for current-accounttransactions only, it becomes easy todisguise capital-account transactions as

14 For further details, see UNCTAD, LegalDimensions of Electronic Commerce, May 4,1999, TD/B/Com.3/EM.8/2.

current-account transactions. This maybe even easier when the purchase isthat of services rather than goods.

Ready access to foreign exchangeis not a problem, however, in so far asexports are concerned. Normally, ex-ports require receipt of foreign ex-change for which restrictions on elec-tronic transmission are likely to be lessof a problem. Moreover, exports arelikely to be undertaken almost exclu-sively by commercial entities whichare generally equipped to deal in for-eign exchange rather than individuals.Even if they need to import certainproducts, they are likely to be able tomake payments electronically in coun-tries with current-account convertibil-ity.

The final step in ensuring access tointernational e-commerce is to haveaccess to communication networks andmarkets for electronically tradablegoods in foreign countries. Access tocommunication networks is essentiallyguaranteed under GATS and theAgreement on Basic Telecommunica-tions as discussed in Section 5.1 of thispaper. At present, there is sufficientexcess capacity in the networks in de-veloped countries. Therefore, access isunlikely to be a problem. It is possiblehowever, that as the use of Internetgrows worldwide, the expansion of ca-pacity may fail to keep up with de-mand. Normally, one would expectthe price mechanism to manage thedemand for access, but there may bephases when networks begin to congestheavily. Under such circumstances,developing countries will need to en-sure that their access rights are notviolated. While, personally, I do notexpect this to become a serious prob-lem, some caution in this regard mayprove valuable.

22

The more important access issuerelates to liberalization commitmentsby developed countries in the servicesthat developing countries can exportelectronically. To-date, liberalizationcommitments by both developed anddeveloping countries have been con-centrated in services traded by mode 3.Developing countries are largely im-porters of these services. Commit-ments in electronically traded services,which developing countries can poten-tially export have been limited.

For some developing countries,the potential for exports of servicesthrough electronic means is very sub-stantial. For instance, the market forcustomized software alone is growingat more than 20 per cent annually andis projected to reach US$ 250 billionby the year 2000.15 Back office serv-ices offer another area in which deve l-oping countries can and have beensupplying services to developed coun-tries. Starting with simple data entryservices in the 1980s, the supply ofback office services from developingcountries has grown to include elec-tronic publishing, website design andmanagement, customer call centres,medical records management, hotelreservations, credit card authorizations,remote secretarial services, mailing listmanagement, technical on-line support,indexing and abstracting services, re-search and technical writing, and tech-nical transcription.

As reported to UNCTAD (1998),and based on OECD (1997) estimates,the global market for back office serv-ices (including Y2K code conversion)that can be potentially supplied by de-

15 The information in this and the follo w-ing paragraph is taken from UNCTAD, July27, 1998, Scope for Expanding Exports of De-veloping Countries in Specific Service Sectors,TD/B/com.1/21.

veloping countries amounted to asmuch as US$ 438 billion in 1998.16

This figure is at least 20 per cent of thetotal 1996 exports of developing coun-tries. United States corporations alonespend US$ 50 billion a year on infor-mation processing, of which at least 20per cent can be provided in a back of-fice environment.

Developing countries should alsoidentify sectors that have not been lib-eralized so far by developed countriesand to which they could export serv-ices electronically. One such areawould seem to be accountancy serv-ices. Negotiations in this are couldpotentially be extremely beneficial tosome of the developing countries sincethis is a very large market.

Internet also offers developingcountries the opportunity to becomeexporters of products purchased byforeign governments. In the past, itwould have been difficult for potentialdeveloping country suppliers to findinformation on these purchases. How-ever, many developed country gov-ernments are now beginning to posttenders for procurement of goods andservices on Internet. This gives sup-pliers from developing countries betteraccess to yet another sector in deve l-oped countries. Though the establish-ment of credibility may take some timefor the small and medium firms, largefirms in developing countries can cer-tainly bid and compete successfully forthese contracts.

16 OECD, 1997, The World in 2020: To-wards a New Global Age.

23

VII. CONCLUSIONS

In this paper, I have discussed themain economic issues relating to e-commerce from the viewpoint of de-veloping countries. The first set of is-sues discussed in the paper concernsthe WTO discipline on this trade. Sev-eral points can be made in this context.First, all things considered, it will bemost appropriate to classify e-commerce as trade in services withGATS discipline applied to it. Sincethis matter is still under negotiation,developing countries should be surethat e-commerce is not classified asgoods trade with a zero customs dutypact made permanent. Such an out-come would liberalize all e-commerceby default, undermining their bargain-ing power.

Second, at present there is somedisagreement about whether the Inter-net transactions in which the providerand recipient of a service are located indifferent countries should be classifiedas cross-border trade or consumptionabroad. In making their commitmentsin the UR and post-UR negotiations inservices, countries presumably viewedthese transactions as cross-bordertrade. For if they are defined as con-sumption abroad, the category de-scribed as cross-border trade in serv-ices will be virtually vacuous. In viewof this fact, it can be argued that thetransactions under consideration beclassified as cross-border trade.

Third, in the area of intellectualproperty protection, developing coun-tries must eventually confront the pos-sibility of two WIPO treaties, con-cluded in December 1996 but yet tocome into force, being brought into theWTO. These treaties have strong en-forcement commitments that develop-

ing countries will need to study care-fully. Many of the countries may lackthe ability to enforce and deliver thesettlement of disputes in this area.

Developing countries such as Indiathat have the capacity to export skilledservices through Internet should ag-gressively negotiate market accesswith developed countries in the forth-coming round. This involves negotia-tions on two fronts. One, they shouldseek liberalization by developed coun-tries in sectors in which they have acomparative advantage. Secondly,they should seek recognition of theireducation, qualifications, requirementsmet, or licences or certificates grantedin the markets of other countries.

Policy issues confronting deve l-oping countries in e-commerce are notlimited to negotiating issues, however.Indeed, for most developing countries,the binding constraints on the deve l-opment of e-commerce are internal.These countries lack adequate tele-communications facilities with thedensity of telephone lines being lessthan three per one hundred people. E-commerce can, of course, grow rapidlyeven when this density is low as theIndian experience testifies. But suchgrowth is likely to be confined to anenclave and will fail to achieve its fullpotential. It can be argued that withsuperior telecommunications infra-structure and regular power supply,even Indian software exports couldhave grown at a much faster pace thanthey did. Efficiency considerationsdictate that, assuming e-commercelowers the costs of transactions, its ex-pansion should not be confined to ex-ternal trade but also extended to do-mestic trade. That, in turn, requires an

24

expansion of telecommunications fa-cilities. Also critical to the expansionof both internal and external e-commerce are financial sector reforms.In particular, unless electronic meansof payment such as credit cards are de-veloped, the expansion of e-commercewill be slow.

Electronic commerce offers un-precedented opportunities to both de-veloping and developed countries.

In the short run, the gains are likely tobe concentrated in developed countriesbut, in the long run, developing coun-tries have more to gain. This is be-cause, in the short run, developingcountries lack the infrastructure neces-sary to take full advantage of Internet.But in the long run, they can leap-frog,skipping some of the stages in the de-velopment of information technologythrough which developed countrieshave had to pass.

UNCTAD Study Series on

POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES

No. 1 Erich Supper, Is there effectively a level playing field for devel-oping country exports?, forthcoming.

No. 3 Joseph Francois, Assessing the results of general equilibriumstudies of multilateral trade negotiations , 2000.

No. 4 John Whalley, What can the developing countries infer fromthe Uruguay Round models for future negotiations?, 2000.

No. 5 Bijit Bora, Peter J. Lloyd, Mari Pangestu, Industrial policy andthe WTO, forthcoming.