trade in services: using openness to grow

Services are vital for economicdevelopment—Services are the fastest growing sector of theglobal economy, and trade and foreign directinvestment in services have grown faster thanin goods over the past decade. Developingcountries have witnessed even faster growthrates, and their share in world services exportsincreased from 14 percent in 1985–89 to 18percent in 1995–98. Technological progresshas greatly enhanced the scope for trade inconventional services, such as education and fi-nance, and also created a host of new tradableservices, such as software development and In-ternet access. Moreover, liberalization in manycountries is leading for the first time to the pri-vate and foreign provision of services such astelecommunications, transport, and finance.

In virtually every country, the performanceof the services sectors can make the differencebetween rapid and sluggish growth (box 3.1).Developing countries, in particular, are likelyto benefit significantly from further domesticliberalization and the elimination of barriersto their exports. The income gains from a re-duction in protection to services are estimatedto be multiples of those from trade liberaliza-tion in goods.

—but the benefits from liberalization arenot automaticFlawed reform programs can substantially re-duce gains. The largest gains come from elim-inating barriers to entry, but many developing

countries have given priority to a change inownership through privatization, while retain-ing limitations on new entry. Effective regula-tion ranging from prudential regulation of fi-nancial services to procompetitive regulationof telecommunications is critical to the successof liberalization, but regulatory weaknessesare too prevalent in developing economies.Liberalization also frequently requires com-plementary policies to help improve access toessential services for the poor. The experienceof several countries has demonstrated thatuniversal service goals can be achieved in com-petitive markets.

Multilateral engagement can be animportant catalyst for liberalizationEven though governments can initiate reformsof services individually, multilateral engage-ment can help in two ways. First, negotiationsunder the General Agreement on Trade in Ser-vices (GATS) could help accelerate domesticreform and improve access to foreign marketsfor developing countries. However, for thesenegotiations to be fruitful, both developed anddeveloping countries must recognize mutualinterests in reciprocal liberalization. In particu-lar, developed countries must see the advan-tages of allowing the temporary movement ofindividual service providers. Developing coun-tries must see the advantages of multilateralagreements to increase competition, enhancecredibility of potential domestic reform, andstrengthen domestic regulation. Recognizing

69

Trade in Services:Using Openness to Grow

3

these potential mutual gains will allow recip-rocal “concessions” that benefit both.

In parallel, global cooperation is needed toprovide support for developing countries atfour levels: in devising sound policies, strength-ening the domestic regulatory environment,enhancing their participation in the develop-ment of international standards, and in ensur-ing access to essential services in the poorestareas.

Surging trade and investment in services

Trade in services: four modes of supplyServices include activities as disparate as trans-port of goods and people, financial interme-diation, communications, distribution, hotelsand restaurants, education, health care, con-struction, and accounting. In contrast to mer-

chandise trade, services are often intangible,invisible, and perishable, and usually requiresimultaneous production and consumption.1

The need in many cases for proximity betweenthe consumer and the producer implies thatone of them must move to make an interna-tional transaction possible. Since the conven-tional definition of trade—where a productcrosses the frontier—would miss out on a wholerange of international transactions, it is nowcustomary to define “trade in services” to in-clude four modes of supply:

• Mode one: cross-border supply, which isanalogous to trade in goods; arises whena service crosses a national frontier, forexample, the purchase of software or in-surance by a consumer from a supplierlocated abroad.

• Mode two: consumption abroad; occurswhen the consumer travels to the terri-

G L O B A L E C O N O M I C P R O S P E C T S

70

In developing countries, the average share of ser-vices in GDP increased from around 40 percent in

1965 to around 50 percent in 1999, while in theOECD countries the average share increased over thesame period from 54 percent to over 60 percent.Among the fastest growing sectors in many countriesare services such as telecommunications, software,and finance.

Efficient services not only provide a direct bene-fit to consumers, but also help shape overall eco-nomic performance. An efficient and well-regulatedfinancial sector leads to the efficient transformationof savings to investment, ensuring that resources aredeployed wherever they have the highest returns; andfacilitates better risk-sharing in the economy. Im-proved efficiency in telecommunications generateseconomywide benefits, because this service is a vitalintermediate input and also crucial to the dissemina-tion and diffusion of knowledge. The spread of theInternet and the dynamism that it has lent toeconomies around the world is telling testimony tothe importance of telecommunications services. Simi-

Box 3.1 Why do services matter for development?larly, transport services contribute to the efficient dis-tribution of goods within a country, and are particu-larly important in influencing a country’s ability toparticipate in global trade. Although these are themore prominent services, others are also crucial.Business services such as accounting and legal ser-vices are important in reducing transaction costs—the high level of which is considered one of the mostsignificant impediments to economic growth inAfrica. Education and health services are necessaryin building up the stock of human capital. Retail andwholesale services are a vital link between producersand consumers, and influence the efficiency withwhich resources are allocated to meet consumerneeds. Software development is the foundation of themodern knowledge-based economy. Environmentalservices contribute to sustainable development byhelping alleviate the negative impact of economic ac-tivity on the environment.

Source: World Bank staff.

tory of service supplier, for example, topurchase tourism, education, or healthservices.

• Mode three: commercial presence; in-volves foreign direct investment, for ex-ample, when a foreign bank or telecom-munications firm establishes a branch orsubsidiary in the territory of a country.

• Mode four: movement of individuals; oc-curs when independent service providersor employees of a multinational firm tem-porarily move to another country.2

Services have been among the fastest grow-ing components of world trade over the last15 years. Over the period 1985–99, the com-pound annual growth rate for services exportson a balance-of-payments basis—which coversprimarily cross-border supply and consump-tion abroad—was over 9 percent per annum,compared to 8.2 percent per annum for mer-chandise (figure 3.1 left). As a result, servicestrade more than trebled its size in fifteen years

to $1.2 trillion in the year 1999, and now ac-counts for a quarter of all cross-border trade.3

Developing countries as a group have wit-nessed an even more rapid (nearly four-fold)increase in their services exports, and a conse-quent increase in their share in world servicetrade from 14 percent in 1985–89 to 18 per-cent in 1995–98 (figure 3.1 right). From a re-gional perspective, Europe and Central Asia(ECA) and East Asia and Pacific (EAP) in-creased their services exports by a factor of six;South Asia (SAR) and Latin America and theCaribbean (LAC) kept up with world growth;and Sub-Saharan Africa (SSA) and the MiddleEast and North Africa (MNA) lagged behind.Even so, most trade in services still takes placebetween rich countries.

Over the last two decades, there has been asignificant decline in the relative importance oftransport services in total services exports—from around one-third to around one-fifth oftotal exports—which may in part reflect a de-cline in the relative price of transport services

T R A D E I N S E R V I C E S : U S I N G O P E N N E S S T O G R O W

71

Figure 3.1 Trade in services has grownfaster than trade in goods—

—and developing countries share in world exports have increased, 1986–98

Note: Population estimate from a sample of 100 countries forperiod 1985–98. Figure for 1999 is estimate from 69 countries.World trade defined as (X+|M|)/2Source: IMF BoP Rev. 5, through SIMA; EPPG staffcalculations.

Note: Population estimate from a sample of 100 countries.Source: IMF BoP Rev. 5, through SIMA; EPPG staffcalculations.

(compound growth, 1985=1) Percent

3.5Services

Goods

3.0

2.5

2.0

1.5

1.0

1985 1987 1989 1981 1993 1995 1997 1999

10.0

1986 1990

18.0

20.6

14.0

18.3

13.7

17.518.0

22.0

1995 1998

12.5

15.0

17.5

20.0

22.5

25.0

GoodsServices

(figure 3.2). While the 1980s witnessed agrowth in the relative importance of travel, the1990s witnessed a significant increase in theshare of other commercial services. Detailed in-formation on this last category is not availablefor most countries. Estimates suggest that fi-nancial services are probably the most impor-tant, followed by construction, communications,and computer and information services.

Most FDI in services goes to OECD economies— A large amount of “trade” in services takesplace through an established presence, for ex-ample through foreign direct investment (FDI).The available evidence suggests that commer-cial presence has been the most dynamic modeof services supply in recent years.4 This may re-flect the fact that there has been far greater lib-eralization of foreign investment than of cross-border supply of services, which was eitheralready open or did not witness significant newopening. At the level of individual sectors, de-spite the growing use of information and com-munications technology, commercial presence

is the dominant mode of supply in all sectorsexcept transport, and to a more limited extenttelecommunications.5

—but the growth rates of FDI flows todeveloping countries are higherThe limited evidence available suggests that thebulk of FDI stocks are in the Organisation forEconomic Co-operation Development (OECD)countries (figure 3.3). However, over the pe-riod 1988–97, stocks in developing countrieshave witnessed much faster rates of growth,increasing ten-fold in EAP, seven-fold in SSA,and five-fold in LAC, compared to a three-foldincrease in the OECD.6 The only exceptionsare the three countries in SSA for which dataare available, where the stock declined by ahalf. In all regions except SSA, the services sec-tor now accounts for nearly half of the entireFDI stock—from 1988 levels of less than one-fifth in LAC and less than one-third in SAS.The limited information on sectoral composi-tion of FDI stocks suggests that nearly half thestock in SAS is in financial services, whereasthe stocks in EAP and LAC are more uniformly


72

Figure 3.2 Transport has declined, while“other” services have increased

“Other” services

Note: Population estimate from a sample of 89 countries.Source: IMF BoP Rev. 5, through SIMA; EPPG staffcalculations.

Source: Trade Handbook, based on IMF BoP rev. 5.

Percent of world total services exports

Communications5% Construction

7%Insurance

5%

Financial10%

Personal, cultural,and recreational

3%

Computer andinformation

5%

Royalties andlicense fees

12%

Otherbusinesses

53%

0Transport Travel Other

10

20

30

40

50

198019901998

distributed across finance, transport, storageand communications, hotels and restaurants,real estate, and other business services.

Developing countries are becoming playersin exporting servicesWhile some developing countries are increas-ingly investing in other countries to exportservices—for example, Malaysia in environmen-tal services, and South Africa in telecommunica-tions—most supply services via cross bordersales (for example, data processing), to visitingforeign consumers (for example, tourism), andthrough the movement abroad of individualservices providers (for example, professionalservices). Developments in information andcommunication technology have dramaticallyincreased the scope for cross-border exports ofservices, ranging from software development inthe Philippines to data processing in Barbados.Rough estimates suggest that the size of the po-tential market for developing-country exports oflong-distance services could be in the range of 1to 5 percent of the total employment in servicesin the seven richest economies—implying ex-

ports valued at between $40 billion and $120billion (World Bank 1995). This mode of deliv-ery is still free of explicit barriers, though regu-latory barriers may impede trade (see box 3.2).

One of the most striking recent examples of a developing-country service export successstory is the Indian software industry. Indiansoftware exports grew from $225 million in1992–93 to $1.75 billion in 1997–98 (at anannual growth rate of approximately 50 per-cent).7 A recent report projects annual rev-enues of $87 billion, 2.2 million jobs, and amarket capitalization of $225 billion for theIndian information technology (IT) sector bythe year 2008.8 By the same year, the IT sectorcould account for 35 percent of India’s ex-ports and attract $5 billion of FDI per year.

These figures are not implausible becauseIndia still accounts for only half a percent ofthe world software market, and there are stillwide differences across countries in the cost of software development and support. The av-erage cost per line of code in Germany (themost expensive country) exceeds by more thanfour times that of India (the cheapest country)


73

Figure 3.3 FDI in services is concentratedin the OECD countries—

—but the growth rates are higher formany developing countries

Source: UNCTAD, FDI/TNC database.

Billions of dollars Simple average compound annual growth rates

0Sub-Saharan

AfricaSouthAsia

LatinAmericaand the

Caribbean

East Asiaand

Pacific

OECD

911.4

304.5

8.9

10146.6

9.47.21.2 0.6 0.8

10.39.5

46.4

38.1

28.3

–4.2

15.3

–2.2

31.7

25.7

100

200

300

400

500

600

700

800

900

1,000

0

Sub-SaharanAfrica

LatinAmericaand the

Caribbean

SouthAsia

East Asiaand

Pacific

OECD–10

10

20

30

40

5019881997

ServicesAll industry


74

Domestic regulations that affect trade pose themain challenge to ensuring open conditions for

electronic delivery of services. Two examples illus-trate how difficult it is to distinguish between regula-tions that incidentally impede trade in the pursuit oflegitimate objectives and regulations that deliberatelydiscriminate against foreign provision for the sake ofprotection.

PrivacyAn issue that could have a profound effect on elec-tronic commerce is privacy. In late 1998, the Euro-pean Union issued a wide-ranging directive that aimsto safeguard the privacy of personal data of EU citi-zens and prevent its misuse worldwide. It is backedby the power to cut off data flows to countries thatthe EU judges not to have adequate data protectionrules and enforcement. The directive caused frictionswith the US, which accused the EU of trying to im-pose laws beyond its own frontiers. A compromisewas reached under which the US agreed to set up ar-rangements for the processing by companies ofpersonal data from the EU, but the issue has notbeen fully resolved.

The issue could have an impact on developingcountries exports of data processing services, andposes a difficult choice for these countries. If theychoose not to enact laws deemed adequate, theycould be shut off from participation in this growingmarket. In the absence of such laws and given theweakness of local legal systems, it might be difficultfor private firms in developing countries to emulateUnited States firms like Microsoft and credibly com-mit to meet the required high standards.

If they do enact stringent laws, it is unlikely thatthey could be made specific to trade with particularjurisdictions, and so the result could be an economy-wide increase in the costs of doing business. For in-stance, if private sector estimates generated in theUnited States are to be believed, information sharingsaves the customers of 90 financial institutions(accounting for 30 percent of industry revenues),$17 billion a year ($195 per average customerhousehold) and 320 million hours annually (4 hoursper average customer household) (Glassman, 2000).

Box 3.2 Whose regulations and for what purpose?Challenges in electronic commerce

It is of course true that reporting of personal credithistories is critical to consumer credit, and, even intheory, excessively, strict privacy laws could createsignificant asymmetries of information and affect theefficiency of markets (Kitchenman, 1999). This isnot to suggest that there might not be good reasonsto protect privacy. However, achieving diverse na-tional objectives without creating unnecessary im-pediments to trade is ideally accomplished through amultilateral process in which developing countriesparticipate.

Offshore financial servicesSeveral Caribbean countries have become off-shorefinancial services centers. However, in recent years,their tax and regulatory regimes have drawn fire andelicited increased scrutiny. For example, the FinancialStability Forum (FSF), which assesses conformity withinternational regulatory standards (including cross-border cooperation) placed many of the Caribbeanoffshore centers in the lowest category; the FinancialAction Task Force (FATF), which is concerned withprotecting financial systems from money launderingand criminal use, placed a number of Caribbean cen-ters in its list of “non-cooperative jurisdictions,” fromthe standpoint of willingness to cooperate with theFATF on the basis of a list of its own criteria; and thecountries also attracted the attention of the OECDfor tax practices deemed harmful.

While the regulatory objectives are legitimate,several concerns have been raised about these initia-tives. First, most developing countries have not par-ticipated in the development of the standards thatare being applied. Second, the standards are not al-ways applied uniformly. For example, the FATF ap-plies the FATF 40 Criteria when conducting mutualevaluations of its members, but uses a different stan-dard, the FATF 25 Criteria, to assess jurisdictionsthat are not FATF members. Third, in some cases theassessment processes are not transparent. For exam-ple, the FSF does not specify how a country classifiedin a low category can improve standards and gradu-ate to a higher category. And FATF deliberations de-termining “non-cooperative jurisdictions” are held inclosed sessions. Finally, the evaluation processes are

(figure 3.4). Against the background of a totalmarket for software services worth about $58billion in the United States, $42 billion in Eu-rope, and $10 billion in Japan, cost savingscould well be substantial.9 Other gains fromtrade liberalization include a more competitivemarket structure for software services, increasedchoice (because countries may develop a specialexpertise for certain development or supportservices), and greater diffusion of knowledge.

The movement of service-supplying person-nel remains a crucial means of delivery. Eventhough the share of on-shore services in totalIndian software exports has been in continu-ous decline (in 1988, the percentage of on-sitedevelopment was almost as high as 90 per-cent), about 60 percent of Indian exports arestill supplied through the temporary movementof programmers to the client’s site overseas.10

Barriers to mode four deprive both homeand host country of benefits Many more developing countries could “ex-port” at least the significant labor componentof services such as construction, distribution,environmental, and transport with greater lib-eralization in the movement of individuals(mode four). If the movement is temporary,then we can be fairly confident that both thehost and home country will gain. For export-ing countries, it is clear that both the financialand knowledge benefits would be greatest ifservice suppliers return home after a certainperiod abroad.11 For importing countries, such

temporary movement should create fewer so-cial and political problems than immigration.

Today, many different barriers constrain themovement of individuals. The most obviousbarriers are explicit quotas or economic needstests—for example, requirements that employ-ers take timely and significant steps to recruitand retain sufficient national workers in thespecialty occupation and that no worker hasbeen laid off for a certain period preceding andfollowing the filing of any work permit or visaapplication.12 Then the many formalities (for


75

Figure 3.4 Software is cheaper todevelop in IndiaCost per line of code (dollars)

Source: Adapted from Rubin (1999).

$25

$15

$10

$5

$0India

$5

$10 $10

$18

$22

$.30 $.60 $.60 $1.10 $1.30

Italy Ireland UnitedStates

Germany

$20

DevelopmentSupport

in some instances not voluntary and involve a “nameand shame” approach to induce compliance.

These issues have provoked continuing discus-sions in the international financial institutions andother fora, but much work needs to be done beforeinternational consensus can be established. The Bankand the Fund are assisting many jurisdictions to as-sess their compliance with international standards

Box 3.2 (continued)

with the aim to help them address any underlyingweaknesses. Key in this is the Bank-Fund Compre-hensive Financial Sector Assessment Programs andthe recent IMF-led program of voluntary off-shore fi-nancial center assessments. Several Caribbean off-shore financial centers have endorsed these initiatives.

Source: Bank staff.

example, to obtain a visa) make red tape re-lated to FDI seem trivial by comparison. Theentry of foreigners can be impeded by non-recognition of their professional qualifications,burdensome licensing requirements, or by theimposition of discriminatory standards on them.The requirement of registration with, or mem-bership of, professional organizations can alsoconstitute an obstacle for a person wishing toprovide the service on a temporary basis.

Health services could be an area ofcomparative advantage—Health services are another area in which de-veloping countries could become major ex-porters, either by attracting foreign patients todomestic hospitals and doctors, or by tem-porarily sending their health personnel abroad.In Cuba, the government’s strategy is to convertCuba into a world medical power. SERVIMED,a trading company created by the government,prepares health and tourism packages. During1995–96 25,000 patients and 1,500 studentswent to Cuba for treatment and training re-spectively, and income earned from sales ofhealth services to foreigners was $25 million.Cost savings for patients and health insurerscan be significant. For example, the cost ofcoronary bypass surgery could be as low as70,000 to 100,000 Indian rupees in India, about5 percent of the cost in developed countries.Similarly, the cost of a liver transplant in India isone-tenth of that in the United States (UNCTADand WHO 1998).

—but will require greater portability ofinsuranceA major barrier to consumption abroad (modetwo) of medical services is the lack of portabilityof health insurance. For example, U.S. federal orstate government reimbursement of medical ex-penses is limited to licensed, certified facilities inthe United States or in a specific U.S. state. Thelack of long-term portability of health coveragefor retirees from OECD countries is also one ofthe major constraints to trade. In the UnitedStates for instance, Medicare covers virtually noservices delivered abroad. Other nations may

extend coverage abroad, but only for limited pe-riods (two or three months). This constraint issignificant because it tends to deter some elderlypersons from traveling or retiring abroad. Thosewho do retire abroad are often forced to returnhome to obtain affordable medical care. If indi-vidual concerns about the quality of care re-ceived abroad are addressed, then the potentialimpact of permitting portability could be sub-stantial. If only 3 percent of the 100 million el-derly persons living in OECD countries retiredto developing countries, they could bring withthem possibly $30 to $50 billion annually inpersonal consumption and $10 to $15 billion in medical expenditures (UNCTAD and WHO1998).

Service reforms can promoteefficiency and growth

Liberalization of trade in services, accom-panied by the reform of complementary

policies, can lead to sectoral and economy-wide improvements in performance.

At the sectoral level—Removing barriers to trade in services in a par-ticular sector is likely to lead to lower prices,improved quality, and greater variety. As in thecase of trade in goods, restrictions on tradereduce welfare because they create a wedgebetween domestic and foreign prices, leading toa loss to consumers that is greater than theincrease in producers’ surplus and governmentrevenue.13 Several empirical sectoral studiessupport this contention.14 Because many ser-vices are inputs into production, the inefficientsupply of such services acts as a tax on produc-tion and prevents the realization of significantgains in productivity. As countries reduce tariffsand other barriers to trade, effective rates ofprotection for manufacturing industries may be-come negative if they continue to be confrontedwith input prices that are higher than theywould be if services markets were competitive.15

A major benefit of liberalization is likely tobe access to a wider variety of services whoseproduction is subject to economies of scale.


76

Consumers derive not only a direct benefit fromdiversity in services such as restaurants and en-tertainment, but also an indirect benefit becausea wider variety of more specialized producerservices, such as telecommunications and fi-nance, can lower the costs of both goods andservices production (Ethier 1982; Copeland2001). In such circumstances, smaller marketscan be shown to have a strong interest in liber-alizing trade in producer services, since this canoffset some of the incentives that firms have tolocate in larger markets (Markusen 1989).16

—and economywide—Estimates of benefits vary for individual coun-tries—from under 1 percent to over 50 percentof gross domestic product (GDP)—dependingon the initial levels of protection and the as-sumed reduction in barriers.17 In simulations ofglobal service trade liberalization, developedcountries gain more in absolute terms—whichis not surprising given the relative size of theireconomies—but developing countries also seesignificant increases in their GDP. One modelpredicts gains of between 1.6 percent of GDP(for India) to 4.2 percent of GDP (for Thailand)if tariff-equivalents of protection were cut byone-third in all countries (Chadha and others2000). The gains from liberalizing services maybe substantially greater than those from liberal-izing trade in goods (box 3.3), because currentlevels of protection are higher and because lib-eralization would also create spillover benefitsfrom the required movement of capital andlabor. For instance, one model finds that thewelfare gains from a 50 percent cut in servicessector protection would be five times largerthan those from nonservices sector trade liber-alization (Robinson and others 1999). Theseresults are particularly striking because they arederived from models that do not fully allow forthe temporary movement of individual servicesuppliers—potentially a major source of gain.

—with accelerator effects on growthCertain services industries clearly possessgrowth-generating characteristics (see box 3.1).Furthermore, barriers to entry in a number of

services sectors, ranging from telecommunica-tions to professional services, are maintainednot only against foreign suppliers but alsoagainst new domestic suppliers. Full liberal-ization can, therefore, lead to enhanced com-petition from both domestic and foreign sup-pliers. Greater foreign factor participation andincreased competition together imply a largerscale of activity, and hence greater scope forgenerating the special growth-enhancing ef-fects.18 Even without scale effects, the importof foreign factors that characterizes servicessector liberalization could still have positiveeffects because they are likely to bring tech-nology with them.19 If greater technologytransfer accompanies services liberalization—either embodied in foreign direct investmentor disembodied—the growth effect will bestronger.20

Econometric evidence—relatively strong forthe financial sector and less strong but nev-ertheless statistically significant for the tele-communications sector—suggests that opennessin services influences long-run growth perfor-mance (figures 3.5 and 3.6). After controllingfor other determinants of growth, countries thatfully liberalized the financial services sectorgrew, on average, about 1 percentage pointfaster than other countries. An even greater im-petus on growth was found to come from fullyliberalizing both the telecommunications andthe financial services sectors. Estimates suggestthat countries that fully liberalized both sectorsgrew, on average, about 1.5 percentage pointsfaster than other countries. While these esti-mates indicate that there are substantial gainsfrom liberalizing key services sectors, it wouldbe wrong to infer that these gains can be real-ized by a mechanical opening up of servicesmarkets.

A flawed reform program can underminethe benefits of liberalizationIf privatization of state monopolies to privateowners (sometimes foreigners) is conductedwithout concern for creating conditions ofcompetition, the result may be merely transfersof monopoly rents to private owners. Similarly,


77

if increased entry into financial sectors is notaccompanied by adequate prudential supervi-sion and full competition, insider-lending andpoor investment decisions may result. Also, ifpolicies to ensure universal service are not put

in place, liberalization need not improve accessto essential services for the poor. Managing re-forms of services markets therefore requires in-tegrating trade opening with a careful combi-nation of competition and regulation.


78

The implications of services liberalization for theTunisian economy have been analyzed by Konan

and Maskus (2000) using a computable general equi-librium model. Using actual data as the foundation,they analyze the effect of liberalizing six service sec-tors: communications, construction, transportation,business and insurance, distribution, and finance.The Tunisian economy is relatively closed, and alsofaces constraints on its exports through the move-ment of individuals. The model is developed so as toconsider three different modes of liberalization: “im-port” liberalization of cross-border trade and theright of establishment by foreign investors, as well asincreased “exports” through cross-border movementof natural persons.

The main finding is that services liberalizationcould provide significant gains to Tunisia, with wel-fare gains equivalent to 7 percent of GDP. These aretwice as large as the gains the model predicts forTunisia from its preferential agreement with the EU.The largest benefits come from the liberalization offoreign investment in financial services, communica-tions, and transportation. Liberalization vitalizes theeconomy by eliminating inefficiency through in-creased international competition. Services are avail-able not only at lower prices but also in greater vari-eties through an increase in the number of firms thatwould operate in Tunisia. More efficient financial,communications, and transportation sectors are alsolikely to attract foreign firms to other industries inTunisia. As more and more foreign firms start to op-erate in Tunisia, the number of varieties of goodsand services made available to consumers and pro-ducers also increases, which further improves wel-fare. The possible cost in terms of restructuring theeconomy turns out to be small. For example, it ispredicted that a mere 3 percent of the workforcewould have to change sectors—a much lower figure

Box 3.3 Welfare gains from service liberalization:The case of Tunisia

than the 6.6 percent adjustment the model predictsas a consequence of the Tunisia-EU free trade agree-ment on goods trade. The gentler impact on thelabor market is a consequence of the fact that ser-vices liberalization induces foreign investment, sothat workers simply change employers within thesame sector. Finally, if Tunisia were to obtain a 20 percent increase in overseas permits for its guestworkers in foreign markets, then there would be anadditional gain in welfare equivalent to 0.4 percentof GDP.

Source: Konan and Maskus 2000.

Percentage change in GDP resulting fromliberalization of selected service sectors

Source: Konan and Maskus 2000.

0.0

Comm

unica

tions

Constr

uctio

n

Trans

porta

tion

Busine

ss

Distrib

ution

Financ

e

0.5

1.0

1.5

2.0

2.5

3.0

South Africa’s experience with liberalizingtelecommunications services is instructive. Thegovernment recognized the need for a moreefficient supply of services. It decided to sell a30 percent equity stake of the public incum-bent, Telkom, to a strategic investor and togrant the newly privatized entity a five-yearmonopoly period for fixed-line telephone ser-vices. It was hoped that market exclusivitywould facilitate rapid infrastructure rollout topreviously underserviced areas, but the pro-gram has had mixed results. Even though net-work growth picked up, Telkom did not meetits rollout obligations and sought to renegoti-ate the targets specified in its monopoly li-cense. The cost of the fixed-line monopolywas also reflected in Telkom’s rising price-costmargin, with gains in productivity leading tohigher margins rather than lower prices (Hodge1999). Finally, despite some improvement, la-


79

Figure 3.5 Services liberalization indices: telecoms & financial services

Note: The openness index for telecommunications captures the degree of competition, restrictions on ownership and theexistence of an independent regulator (needed to enable competitive entry), and draws on an ITU-World Bank database for 1998.The index for financial services captures the restrictions on new entry, foreign ownership and capital mobility, and draws primarilyupon commitments made by countries under the GATS, which are known to reflect closely actual policy, and data in the IMF’sAnnual Report on Exchange Arrangements and Exchange Restrictions.Source: Mattoo, Rathindran, and Subramanian 2001.

South Asia (3)

East Asia and Pacific (5)FinancialServices

Telecoms

0.0

Less restrictive

2.0 4.0 6.0 10.08.0

Eastern Europe and Central Asia (3)

South Asia (5)

East Asia and Pacific (8)

Latin America and Caribbean (21)

High Income (21)

Eastern Europe and Central Asia (3)

Latin America and Caribbean (18)

High Income (26)

Sub-Saharan Africa/Middle East and North Africa (17)

Sub-Saharan Africa/Middle East and North Africa (42)

�

Figure 3.6 Greater liberalization in services is associated with more rapid growthGrowth rate (adjusted for other factors)

Source: Mattoo, Rathindran, and Subramanian 2001.

–.024

Composite services liberalization1 8.5

.059

Linear prediction

ITA

NZL

SLV

PAN

PRT

FIN

ISL

ARG

CHE

ESP

BEL NLD

NOR USA

CAN SGP

EGY

FRA

SWE

GRC

GUY

AUS

GBR

AUT

CYP JAM

BOL

DNK

MLT

MWI

CRI

ZAF

TUR

MOZ

KEN

IND

MAR

VEN

MEX

PHL

NIC CHL

URY KOR

PER

MYS

ECU

AGO

THA

HND

COL

TUN LKA

BRA

IDN

DOM

bor productivity was only a quarter that of lead-ing international operators, with the lack ofcompetition in the domestic market identifiedas a major contributing factor. Continued re-strictions on domestic and foreign entry ap-pear to have prevented the realization of thefull benefits of competitive markets.

In addition to competition, the institutionaland regulatory framework plays a critical role.For example, in the 1990s financial reformswere introduced in many African countries,but have been less successful than expected(World Bank 2001a). Some of the reasons forthe disappointing results are directly related to the financial system, while others pertain to the general economic environment. The re-structuring of state-owned banks was not suf-ficient to change the behavior of the financialinstitutions. Public authorities still pressuredthese institutions to lend money to loss-mak-ing public enterprises. Liberalization failed totrigger competition in the banking sector andgovernments were generally reluctant to closedown distressed state banks. Furthermore, lib-eralization of interest rates in a setting charac-terized by uncontrolled fiscal deficits had a per-nicious effect on domestic public debt, which inturn led to larger deficits. Finally, and crucially,there was a lack of adequate regulation and su-pervision mechanisms to monitor the function-ing of the financial system.

The collapse of the Republic of Korea’s econ-omy in 1997 also reveals the precariousness offinancial liberalization in an imperfect policyenvironment. Korea did liberalize its financialmarkets substantially, but it encouraged thedevelopment of a highly fragile financial struc-ture.21 By liberalizing short-term foreign bor-rowing, the Korean authorities made it possi-ble for the larger and better-known banks andconglomerates (chaebols) to assume heavy in-debtedness in short-term foreign currency debt.Meanwhile, the second tier of large chaebolsgreatly increased their short-term indebtednessin the domestic financial markets (funded indi-rectly through foreign borrowing of the banks).The funds borrowed were being invested in theover-expansion of productive capacity. At the

same time, financial regulation and supervisionwere fragmented with responsibilities spreadunclearly between the Bank of Korea and sev-eral parts of the Ministry of Finance. In addi-tion, Korea had a restrictive regime in terms offoreign bank entry. Until the 1997 crisis, theKorean banking system was virtually closed toforeign banks, in contrast to some other EastAsian economies, such as Hong Kong (China),which was almost completely open for all fi-nancial services. This restrictive regime impededthe development of the local institutions, andmay have contributed to the large capital out-flows as foreign creditors refused to rollovertheir loans.

Liberalization could increase prices ofsome services for the poor—Opening up essential services to foreign or do-mestic competition could have an adverse ef-fect on the poor—which is often cited as a rea-son for the persistence of public monopolies.However, a more efficient solution is to haveregulations with a social purpose.

If a country is a relatively inefficient pro-ducer of a service, liberalization and the resul-tant foreign competition are likely to lead to adecline in domestic prices and improvement inquality. But there is a twist. Frequently, theprices before liberalization are not determinedby the market but set administratively, and arekept artificially low for certain categories ofend-users or types of services products. Thusrural borrowers may pay lower interest ratesthan urban borrowers, and prices of local tele-phone calls and public transport may be keptlower than the cost of provision.22 This struc-ture of prices is often sustained through cross-subsidization within public monopolies, orthrough government financial support.

Liberalization threatens these arrangements.Elimination of restrictions on entry imply anend to cross-subsidization, because it is nolonger possible for firms to make extranormalprofits in certain market segments. New en-trants may focus on the most profitable mar-ket segments (“cream-skimming”), such asurban areas, where network costs are lower


80

and incomes higher. And privatization couldmean the end of government support. The re-sult is that even though the sector becomesmore efficient and average prices decline, theprices for certain end-users may actually in-crease or availability decline, or both.

The evidence on the relationship betweencompetitive market structures and wider ac-cess to services is mixed. In some cases, a pos-itive relationship has been observed in servicessuch as basic telecommunications, especiallyin countries where initial conditions are fee-ble, as exemplified by a low teledensity or ser-vice rationing (long waiting lists for obtainingconnections). However, there is also evidencethat financial services liberalization in somecountries has had an adverse affect on accessto credit for rural areas and the poor.23 Thesepoint to the need to create mechanisms to en-sure that the poor have adequate access to ser-vices in liberalized markets.

—and entail adjustment costs Different modes of supply have different ef-fects on factor markets. Cross-border tradeand consumption abroad resemble goods tradein their implications. The impact of the move-ment of factors depends critically on whetherthe factors are substitutes or complements fordomestic factor services. Given the structure offactor prices in poor countries, we would typi-cally expect liberalization to lead to an inflowof capital and skilled workers. Such inflowswould tend to be to the advantage of the un-skilled poor, increasing their employment op-portunities and wages.24 Interestingly, it hasbeen shown that even when foreigners com-pete with local skilled workers in a services sec-tor, the productivity boost to the sector fromallowing foreigners access could lead to an in-crease in the demand for domestic skilledworkers—the scale effect could outweigh thesubstitution effect (Markusen, Rutherford, andTarr 2000). Given these predictions, why areworkers in developing countries sometimesskeptical about the benefits of liberalization?One concern is the possible reduction in em-ployment in formerly public monopolies that

have frequently employed surplus labor. Forexample, Alexander and Estache (1999) findthat the privatization of electricity distributionin Argentina led to a 40 percent reduction inthe workforce after privatization.

But there is also evidence that pessimismmay not always be justified. For example, anumber of developing countries have managedto maintain or even increase employment intheir liberalized telecommunications sectors.Since many developing countries have low tele-densities (in the vicinity of five lines per 100people), roughly 70 percent of telecom invest-ment in developing countries is directed to-ward building wire line and mobile networks,which are labor intensive and hence help main-tain or raise employment levels. Petrazzini andLovelock (1996) find in a study of 26 LatinAmerican and Asian economies that telecommarkets with competition were the only onesthat consistently increased employment levels,while two-thirds of the countries with monop-olies saw considerable declines in their telecomworkforce.25 Nonetheless, reform programs willgenerally require complementary policies to miti-gate any social and economic costs of adjust-ment in factor markets.

Domestic policy: emphasizingcompetition and regulation

Increasing competition is the first order of businessMany developing countries have moved awayfrom public monopolies in sectors such as com-munications, financial, and transport services,but are still reluctant to allow unrestricted newentry. Privatization does not axiomatically meangreater competition. Restrictions on foreignpresence assume particular significance in thecase of services where cross-border delivery isnot possible, because consumer prices then de-pend completely on the domestic market struc-ture. Several studies have concluded that largerwelfare gains arise from an increase in competi-tion than from a simple change in ownershipfrom public to private hands (Armstrong and


81

others 1994). Foreign investment clearly bringsbenefits even in situations where it does not leadto enhanced competition. Foreign equity mayrelax a capital constraint, can help ensure thatweak domestic firms are bolstered (for example,via recapitalizing financial institutions), andserve as a vehicle for transferring technologyand know-how, including improved manage-ment. However, if restrictions on competitionartificially inflate the returns on investment, thenet returns to the host country may be negative.

Are there good reasons to limit entry? Insome cases, technical limitations may preventcompetition—such as those imposed by thescarcity of radio spectrum needed for the pro-vision of mobile telecommunications services,and scarcity of space for department stores orairports in a city. More generally, entry restric-tions might be justified by the existence of sig-nificant economies of scale. For example, ifthere are substantial fixed costs of networks,competitive entry could lead to inefficient net-work duplication.26 However, entry restrictionsare increasingly hard to defend in principle, inthe face of technological change and in the faceof mounting evidence that competition works.

First of all, entry restrictions change the na-ture of interaction between incumbents andmay well make collusion more likely. Second,such restrictions dampen the impact of compe-tition on productive efficiency. Third, the reg-ulator is usually not better placed than thecompetitive process to determine the optimalnumber of firms in the market, especially giventhe difficulty of obtaining information aboutthe cost structure of firms and other sources ofregulatory failure. Furthermore, technologicaladvances have significantly lowered networkcosts in a unisector such as telecommunica-tions, and vertical separation (for example,through network unbundling) has widened thescope for competitive entry (Smith 1995).Therefore inefficiencies introduced by duplica-tion of networks may be small compared tooperational inefficiencies that can result froma lack of competitive pressure.27 For example,even in telecommunications, a sector where

fixed costs are significant, countries in LatinAmerica that granted monopoly privileges totelecom operators of six to ten years to the pri-vatized state enterprises saw connections growat one and a half times the rate achieved understate monopolies, but only half the rate inChile, where the government retained the rightto issue competing licenses at any time (Welle-nius 1997). A recent study of countries in Asiafound that the largest increases in mainlinepenetration and productivity were witnessedin countries where a change of ownership wasaccompanied by the introduction of competi-tion and the strengthening of regulation (Finkand others 2001).

Efficient regulation: Making competition workRegulation in services, as in goods, arises es-sentially from market failure, which is attrib-utable to the problems of natural monopolyand inadequate consumer information, andfrom considerations of equity and protectingthe poor.

The existence of natural monopoly or oli-gopoly is a feature of the so-called locationalservices. Such services require specialized dis-tribution networks: roads and rails for landtransport, cables and satellites for communica-tions, and pipes for sewerage and energy dis-tribution (UNCTAD; and World Bank 1994).

Many countries have instituted indepen-dent regulators for basic telecommunicationsservices to ensure that monopolistic suppliersdo not undermine market access by chargingprohibitive rates for interconnection to theirestablished networks (see box 3.4).28 A simi-lar approach is being taken in a variety ofother network services, including transport(terminals and infrastructure), and energy ser-vices (distribution networks).

Regulation of the interconnection price maynot, however, be sufficient. Small markets maynot be able to create conditions for effectivecompetition in the supplies of certain telecom-munications, transport, and financial services,even if they eliminate all barriers to entry—for


82

two related reasons. First, with services, unlikein the case of goods, national markets are oftensegmented from the international market dueto the infeasibility of cross-border delivery. Sec-

ond, changing technologies may have reducedthe optimal scale of operation as well as sunkcosts in these sectors, but not enough for smallmarkets to sustain competitive market struc-


83

It is now widely recognized that in basic telecom-munications procompetitive regulation is needed to

deliver effective competition and gains from liberal-ization. But the experience of different countries re-veals a range of political and economic difficultiesthat are only gradually being overcome.

In India a conflict between the department of telecommunications (DOT) and the regulatoryagency, Telecommunications Regulatory Authority ofIndia (TRAI), as it was initially constituted, hamperedprogress toward an efficient telecom infrastructure.Underlying a number of these problems was theDOT’s joint role in awarding licenses for both basicand cellular services while remaining as the maintelecommunications service provider. Absent an inde-pendent regulator, empowered to rebalance tariffs,enforce fair interconnection agreements, and ensurerapid, equitable issuance of radio spectrum, the bene-fits of a sector opened to allow private participationand foreign investment were significantly limited.

The government announced a new telecom-munications policy on March 26, 1999 that ad-dressed several of these key outstanding issues. TheDOT’s policymaking and service provision functionswere separated, and the operations arm was corpora-tized. TRAI was reconstituted in 2000, and its dis-pute resolution powers are now vested in a newquasi-judicial agency. The authority announced a newtelephone tariffs decision that will substantially re-structure telephone service prices over a three-yearperiod, significantly improving incentives for localnetwork investment. The regulator has also pro-grammed an agenda of activity to address severalother important regulatory matters, such as intercon-nection arrangements; a numbering plan; quality ofservice; rules of business; and customer satisfaction.

For smaller countries, a different problem arises:the creation and operation of an efficient regulatoryagency involves substantial fixed costs that could

Box 3.4 Challenges in implementing procompetitive regulation

place a significant resource burden. Apart from spec-trum monitoring equipment, computers, and pro-grams, there is the cost of professional assistance foractivities such as interconnection, cost estimation, andspectrum management. For example, the total cost of government in Dominica is $41 million a year,whereas the budget of the U.S. telecom regulator (theFederal Communications Commission) runs to $210 ayear. It is estimated that even a bare-bones regulatoryauthority is likely to cost in the region $2 million eachyear, or 5 percent of Dominica’s government budget.

In response to these problems, in May 2000, St Lucia, Dominica, Grenada, St Vincent and theGrenadines, and St Kitts and Nevis set up, withWorld Bank support, the Eastern Caribbean Tele-communications Authority (ECTEL), the first re-gional telecommunications authority in the world.ECTEL is in the process of developing from a legalentity into a functioning institution. Although themember countries will retain their sovereign powerover licensing and regulation, ECTEL will providetechnical expertise, advice, and support for nationalregulations. Apart from the economies of scale in es-tablishing a common regulator, there are at leastthree other advantages. It will promote the develop-ment of harmonized and transparent regulation inthe region, allow for a greater degree of indepen-dence (and hence credibility) in regulatory advice,and enhance bargaining power in negotiations withincumbents and potential entrants. In fact, there isevidence that the creation of ECTEL, along withother reforms, has already prompted a decline in theprices of telecommunication services in the region.One example is that the per-minute cost of a daytimecall to the United States has fallen between 24 and42 percent in these countries.

Source: DeFreitas, Kenny, and Schware 2001; and World Bankstaff.

tures. Some form of final price regulation may,therefore, be unavoidable. In some cases, suchregulation can be implemented at the nationallevel although, in practice, many developingcountries today lack the means to do so. Inother cases, the limited enforcement capacity ofsmall states strengthens the case for multilat-eral initiatives.29

Regulation to remedy inadequateconsumer information In many intermediation and knowledge-basedservices, consumers have difficulty securing fullinformation about the quality of service they arebuying (UNCTAD and World Bank 1994). Con-sumers cannot easily assess the competence ofprofessionals such as doctors and lawyers, thesafety of transport services, or the soundness ofbanks and insurance companies. When such in-formation is costly to obtain and disseminate,and consumers have similar preferences about therelevant attributes of the service supplier, the reg-ulation of entry and operations in a sector couldincrease social welfare. However, the establish-ment of institutions competent to regulate well isa serious challenge, as is revealed by the difficul-ties in the financial sector—not only in a numberof developing countries but also in the UnitedStates, Sweden, and Finland in the 1980s and1990s. The fact that regulatory inadequacies can-not be quickly remedied raises the issue of howdifferent elements of reform—particularly pru-dential strengthening and trade and investmentliberalization—are best sequenced (see box 3.5).

A separate problem is that domestic regula-tions to deal with the market failure may them-selves become impediments to competition andtrade, as a result of differences across jurisdic-tions in technical standards, prudential regula-tions, and qualification requirements in profes-sional, financial, and numerous other services(see box 3.2). In many cases, the impact ontrade is an incidental consequence of the pur-suit of a legitimate objective, but in some casesregulation can be a particularly attractivemeans of protecting domestic suppliers fromforeign competition.30 The issue of how multi-lateral trade rules might sift the legitimate from

the protectionist is an issue to which we returnin the final section of this chapter.

Regulation to ensure universal service Reform programs can accommodate universalservice obligations by imposing this require-ment on new entrants in a nondiscriminatoryway. Thus such obligations were part of thelicense conditions for new entrants into thefixed network telephony and transport in sev-eral countries. However, subsidies have oftenproved more successful than direct regulationin ensuring universal access (Estache and oth-ers 2001).31 In 1999, Peru adopted a universalservice levy of 1 percent to finance a fund ded-icated to providing universal access in remoteareas. Funds were allocated through a com-petitive bidding process that encouraged oper-ators to adopt the best technology and othercost-saving practices at minimum subsidy. TheChilean government adopted a similar schemethat permitted it to leverage over $2 million inpublic funds into $40 million in private in-vestment; this resulted in installation of tele-phones in 1,000 localities at about 10 percentof the costs of direct public provision. House-hold ownership of a telephone in Chile in-creased from 16 to 74 percent from 1988 to2000, and all but 1 percent of the remaininghouseholds were provided with public accessto telephones.

Public subsidies may also be directed to theconsumer rather than the provider (Cowheyand Klimenko 1999). Governments have ex-perimented with various forms of vouchers,from education to energy services. This last in-strument has at least three advantages: first, itcan be targeted more directly to those who needthe service and cannot afford it; second, itavoids the distortions that arise from artificiallylow pricing of services to ensure access; and fi-nally, it is an instrument that does not discrim-inate in any way between providers. Of course,no single approach will fit all sectors and coun-tries, and the appropriate model to ensure ser-vice delivery to low-income groups will dependon local circumstances, particularly regulatorycapacity.


84


85

Financial reform is especially complicated. It isuseful to distinguish three types of financial liber-

alization and the scope of each.

• Domestic financial liberalization allows marketforces to work by eliminating controls on lendingand deposit rates and on credit allocation and,more generally, by reducing the role of the state inthe domestic financial system.

• Capital account liberalization removes controls on the movement of capital in and out of acountry and restrictions on the convertibility ofcurrency.

• Internationalization of financial services eliminatesdiscrimination in treatment between foreign anddomestic financial services providers, and removesbarriers to the cross-border provision of financialservices.

Internationalization has raised several fears: thethreat to the survival of local banks and financialcompanies; the loss of monetary autonomy; and theincreased volatility of capital flows. Many of theseconcerns do not relate just to internationalization offinancial services, but also to the processes of finan-cial deregulation and capital account liberalization.But the extent of benefits and costs of international-ization depends, to a great extent, on how it isphased in with these other two types of financial re-form, and, in particular, the strengthening of pruden-tial regulation and supervision.

Many countries that have successful experiencesopening up to foreign financial firms (Brazil, Chile,Hungary, Ireland, Poland, Portugal, Spain, and others)also engaged in a process of domestic deregulationand, consequently, reaped substantial gains (WorldBank 2001b). The experience of the countries acced-ing to the EU suggests that internationalization anddomestic deregulation can be mutually reinforcing.Increased foreign entry bolstered the financial sectorframework by creating a constituency for improvedregulation and supervision, better disclosure rules, andimprovements in the legal and regulatory frameworkfor the provision of financial services. It also added tothe credibility of rules. These benefits of opening up toforeign entry followed from both top-down actions onthe part of government, as well as from bottom-up

Box 3.5 Financial sector liberalization:the need for policy coherence

pressures from the market as best international prac-tices and experiences were introduced.

While the two reform processes (international-ization and domestic financial deregulation) are mu-tually reinforcing, they are not sufficient in them-selves. More than in other sectors, the gains andcosts of financial reform depend on the regulatoryand supervisory framework, (Barth and others2001). Experience shows that it is vital to strengthenthe supporting institutional framework in parallelwith domestic deregulation and internationalization.In the absence of such strengthening, foreign entrymay entail risks. Foreign bank entry can destabilizelocal banks by taking away the lowest risk busi-ness—including large, exporting firms—leaving localbanks to venture further out on the risk frontier.Also several countries, especially in Africa, discov-ered with the failure of banks—such as BCCI andMeridien—that a foreign name did not guaranteesafety and soundness even when these foreign bankswere operating in industrial economies or had someownership links with reputable foreign sources.

Having a supportive institutional framework iseven more obvious when it comes to capital accountliberalization. Experiences in the past, most recentlyin Asia, have shown that achieving the potentialgains, and avoiding the risks, of capital account liber-alization depend to a great extent on whether domes-tic institutions and prudential authorities have devel-oped sufficiently to ensure that foreign finance will bechanneled in productive directions (Eichengreen forth-coming). Recent experiences also shows the potentialbenefits of foreign financial institutions in stabilizingcapital flows. Several countries with significant for-eign presence, such as Argentina and Mexico, bene-fited from the access of these institutions to foreigncapital during periods of economic presence (Dages,Goldberg, and Kinney 2000). More generally, studiesshow that diversity in ownership contributes togreater stability of credit in times of crises (Barth andothers 2000a; b); and La Porta and others 2000). Inso far as foreign presence leads to a stronger regula-tory and supervisory framework, it contributes tomaking capital account liberalization and internation-alization mutually reinforcing.

Source: World Bank staff.

Multilateral engagement:Buttressing domestic reforms

In principle, a country can liberalize its mar-kets and strengthen its regulatory institu-

tions unilaterally, but four types of issues cre-ate benefits from multilateral engagement.First, liberalization may be constrained by do-mestic opposition from those who benefit fromprotection. Second, a country cannot on itsown improve access for its exports to foreignmarkets. Third, a small country may not beable to deal adequately with anticompetitivepractices by foreign suppliers. Finally, a coun-try may lack the expertise and resources to de-vise and implement optimal policy, especiallyin the area of domestic regulation.

The WTO is the natural forum to pit thefirst two elements—opposition to reform athome and barriers to access abroad—againsteach other constructively through the processof mercantilist negotiations. But there is also a need for complementary multilateral efforts

to ensure that the gains from liberalization are not undermined by inadequacies in policychoice and regulation.

Using the current round of GATSnegotiations to deliver liberalization athome and access to markets abroadThe General Agreement on Trade in Services(GATS) had a deliberately symmetric struc-ture, encompassing the movement of bothcapital and labor for services provision. In the-ory, developed and developing countries couldindeed bargain to exploit their modal compar-ative advantage: improved access for capitalfrom developed countries being exchanged forimproved temporary access for individual ser-vice providers from developing countries. Inpractice, all countries have been unwilling togrant greater access for foreign individuals(except for the limited class of skilled intra-corporate transferees), and a tradeoff betweenmodes of delivery simply has not occurred


86

Figure 3.7 WTO members have been reluctant to make market access commitments on themovement of natural persons (Mode 4)

0

Note: Calculated on the basis of a sample of 37 sectors deemed representative for various services sectors.Source: See WTO Document S/C/W/99, March 2, 1999.

LegendThe upper part of each bar represents partial commitments, the lower part full commitments.DC = Developed countriesLDC = Developing and transition economiesAC = Acceding countries

DC LDC ACMode 4

DC LDC ACMode 2

DC LDC ACMode 3

DC LDC ACMode 1

10

20

30

40

50

60

70

80

90

100

(figure 3.7). Moreover, even the negotiatinglinks across services sectors and between ser-vices and goods sectors do not seem to havebeen particularly fruitful. And so, since gov-ernments could not demonstrate improved ac-cess to foreign markets as a payoff for domes-tic reform, GATS commitments reflect for themost part the existing levels of unilaterally de-termined policy—rather than liberalizationachieved through a reciprocal exchange of“concessions.”32

This may change with time. With severeshortages of skilled labor in the United Statesand Europe and the powerful constituency ofhigh-technology companies lobbying for re-laxation of visa limits, the prospects for seri-ous intermodal tradeoffs—such as obtainingtemporary labor movement in return for al-lowing greater commercial presence for for-eign service providers—are now greater. Thechallenge is, first, to devise mechanisms thatprovide credible assurance that movement istemporary rather than a stepping-stone to mi-gration; and second, to devise negotiating for-mulae that credibly link Mode 4 liberalizationto reductions in restrictions in other areas.

Strengthening GATS rules and commitmentsIn line with the WTO’s central concern withsecuring market access, it would also be nat-ural to use the GATS to enhance the credibil-ity of policy at home and security of access tomarkets abroad through legally binding com-mitments; to ensure that domestic regulationssupport trade liberalization; and to prevent dis-crimination between trading partners by ensur-ing effective application of the most-favorednation (MFN) principle.33

First, the GATS could help secure access tomarkets that are already open. Trade in elec-tronically delivered products, in which moreand more developing countries are beginningto participate, must continue to remain free of explicit barriers—should such barriers everbecome feasible. It would be far more effectiveto widen and deepen commitments under theGATS on cross-border trade (see box 3.6).

At home, policies that are believed in aremost likely to succeed. Developing countriesthemselves could take greater advantage of theopportunity offered by the GATS to lend credi-bility to reform by committing to maintain cur-rent levels of openness or to greater levels of fu-ture openness. In basic telecommunications, theone sector where countries have been willing tomake such commitments, there is evidence thatthe commitments have facilitated reform.

Developing countries have much to gainfrom stronger multilateral rules on domestic reg-ulations. Such rules can play a role in promotingand consolidating domestic regulatory reform,as happened to some degree in the telecommu-nications negotiations. The rules are also neededto equip developing-country exporters to ad-dress regulatory barriers in foreign markets inthe form of burdensome licensing and qualifica-tion requirements for professionals, or restric-tive standards in electronic commerce.

It is desirable also to remedy the currentweaknesses in the application of the MFNprinciple in the GATS. One obvious problemis the explicit departure from the MFN obli-gation through numerous MFN exemptionslisted by countries. Less visible, but potentiallymore serious, is the possibility of implicit dis-crimination through preferential recognitionagreements and allocation of quotas. Rules inthese areas need to be clarified and strength-ened to protect developing countries bothfrom discrimination in their export marketsand from pressure to grant particular foreignsuppliers privileged access to their markets—as, for instance, is reported to be happening inthe Chinese insurance market.

Dealing with anticompetitive practicesAnticompetitive practices that fall outside thejurisdiction of national competition laws maybe important in sectors such as maritime, airtransport, and communication services. Thecurrent GATS provision in this area providesonly for information exchange and consulta-tion. Strengthened multilateral rules are neededto reassure small countries with weak enforce-ment capacity that the gains from liberaliza-


87


88

Trade in electronically delivered prod-ucts, in which more and more develop-

ing countries are beginning to participate,and which offers an increasingly viablealternative to the movement of individuals,is today largely free of explicit barriers. Themain concern should be preventing the in-troduction of new barriers if they becometechnically feasible. What is the best routeto preventing the imposition of explicit re-strictions, such as tariffs and quotas? (Theissue of regulatory barriers is discussed inbox 3.2.)

WTO Members have so far focused onprohibiting the imposition of customs du-ties on electronically delivered products. Itis ironic that considerable negotiation en-ergy has been invested in prohibiting theeconomically superior (and probably notfeasible) instrument of protection whereaslittle attention has been devoted to inferior(and possibly more feasible) instrumentssuch as quotas and discriminatory internaltaxation. In any case, since the bulk of suchcommerce concerns services, open tradingconditions are more effectively securedthrough deeper and wider commitmentsunder the GATS on cross-border trade re-garding market access (which would pre-clude quantitative restrictions) and nationaltreatment (which would preclude all formsof discriminatory taxation).

There is considerable scope for an im-provement in such commitments. For instance indata processing, of the total WTO Membership ofover 130, only 66 Members have made commit-ments; and only around two-thirds of these commit-ments guarantee unrestricted market access. Manydeveloping countries have not made sectoral commit-ments, but the commitments of the few which have,are frequently superior to those of developed coun-tries. It is particularly striking that in some of thecore financial services, about a third of the develop-ing countries which have made commitments guaran-tee unrestricted cross-border supply, whereas none ofthe 26 developed countries does so. Developingcountries have also been more forthcoming than de-veloped countries in audiovisual and entertainmentservices. One possible approach to improving com-

Box 3.6 Ensuring barrier-free trade in electronicallydelivered products

mitments would be for all Members to agree that norestrictions would be imposed on cross-border deliv-ery, either of all services or of a bundle whose com-position could be negotiated.

These commitments have additional value be-cause other GATS disciplines, for example, on do-mestic regulations, would only meaningfully kick inonce these commitments are in place. For instance, ifthere were excessively restrictive regulatory barriersto cross-border trade in the core banking services indeveloped countries, it would be difficult today tochallenge them, since these countries have not evencommitted to provide market access and nationaltreatment.

Source: Mattoo and Schuknecht 2000.

Commitments on cross-border supply in selectedservices sectors

Source: World Trade Organization.

*Number of countries with commitments.Full Partial

Dataprocessing services

Voicetelephone services

Online info anddatabase retrieval

Audiovisualservices

Retailingservices

Adulteducation

Non-lifeinsurance

Acceptanceof deposits

Lending ofall types

Trading insecurities

Entertainmentservices

News agencyservices

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

26

C

DC

LDC 40

DC

LDC

25

51

DC

LDC

26

39

DC

LDC

4

40

DC

LD

25

19

DC

LDC

18

13

DC

LDC

26

48

DC

LDC

25

55

DC

LDC

25

54

DC

LDC

26

45

DC

LDC

17

22

DC

LDC

22

3

*

tion will not be appropriated by internationalcartels. For instance, the United States and theEU could begin by ending the exemption ofcooperative price-setting and related practicesin maritime transport from the scope of theircompetition law. Ending the exemption wouldenable a careful assessment by competition au-thorities of the social costs and benefits of thesecollusive arrangements. Competitive disciplinecould also be strengthened by creating a rightfor foreign consumers to challenge anticompet-itive practices by services firms in the nationalcourts of countries whose citizens own or con-trol these firms—a variant of the precedent inthe WTO rules on intellectual property and gov-ernment procurement.

Global cooperation to supportliberalizationBeyond WTO negotiations multilateral sup-port is needed at four levels: in devising soundpolicy, strengthening the regulatory environ-ment, enhancing developing country partici-pation in the development of internationalstandards, and ensuring access to essential ser-vices in the poorest areas.

While there is growing consensus on thebenefits of liberalization, there is less agree-ment on the precise route to liberalization.Certain issues have prompted differing strate-gies. Should all barriers to entry be eliminatedin sectors with significant economies of scale?How far should trade and investment liberal-ization be conditioned on strengthened pruden-tial regulation? Developing countries in partic-ular could benefit from the experience of othercountries on these issues—but the experienceswith electricity in California and rail transportin Britain suggest that there is scope for learn-ing in all countries. More work is needed at thenational and international levels to take stockof individual and cross-country experience toidentify the areas where there are clear pre-scriptions for policy and those where there is aneed for further research, and therefore for hu-mility in policy advice and formulation.

Sound domestic regulation—ranging fromprudential regulation in financial and profes-

sional services to procompetitive regulation ina variety of network-based services—is criticalto realizing the benefits of services liberaliza-tion. We have also seen that devising and im-plementing such regulation is not easy, andthat there are acute regulatory problems inmany developing countries. Regulatory insti-tutions can be costly and may require sophis-ticated skills. To some extent such costs can berecovered through fees or regional coopera-tion—but external assistance could help en-sure that adequate regulation is in place. Sometechnical assistance is already being provided,but often on an ad hoc basis either bilaterallyor through international organizations. Moresystematic efforts—along the lines of the Inte-grated Framework for least-developed coun-tries—are needed to assess the needs of indi-vidual developing countries and to ensure thatthe most appropriate assistance is provided inkey sectors.

Improvements in domestic standards andqualifications are also needed in order to ex-port services. For example, in the case of pro-fessional services, low standards and dispari-ties in domestic training and examinations canbecome a major impediment to obtaining for-eign recognition. Thus inadequacies in domes-tic regulation can legitimize external barriersto trade. At the same time, developing coun-tries need to participate more actively in thedevelopment of international regulations andstandards, especially in new areas such as elec-tronic commerce. Otherwise, standards couldevolve to reflect the concerns only of devel-oped countries and impede the participationof developing countries in services trade.

There will remain certain poor countries, orcertain regions within poor countries, whereimprovements in services policy and regulationwill not be sufficient to ensure access to es-sential services. The criterion for determiningwhether assistance is needed could be the ab-sence of private sector provision despite com-prehensive policy reform. International as-sistance effectiveness could be maximized byallocating it in a manner similar to that useddomestically by countries such as Chile and


89

Peru to achieve universal service. For instance,once a country (or a region within a country)has been selected for assistance, funds—such asthose provided by certain countries to bridgethe digital divide—could be pooled and allo-cated through international competitive tendersto the firm that offers to provide the necessaryinfrastructure at least cost. Providing inter-national assistance in meeting the costs of therequired subsidy programs could increase thebenefits of, and facilitate, liberalization by en-suring that the needs of the poor would be met.

Notes1. There are, however, exceptions to each of these

characteristics of services: a software program on adiskette or an architect’s design on paper are both tan-gible and storable, many artistic performances are vis-ible, and automated cash-dispensing machines makeface-to-face contact between producers and consumersunnecessary. These exceptions do not, however, detractfrom the usefulness of the general definition of servicespresented above.

2. This view of trade originated in Bhagwati 1984and Sampson and Snape 1985, and has been formal-ized in the General Agreement on Trade in Services(GATS).

3. The invisibility and intangibility of most servicesimply that when they are delivered across borders, their passage is not recorded by a customs official. Dataon services are therefore unreliable and volatile. Fur-thermore, statisticians in most countries do not keeptrack of the sales of services by foreign investors or for-eign individuals who stay for longer than a year. Despitethese difficulties, it is possible to put together a roughpicture of trade in services by drawing on three com-plementary sources. The International Monetary Fund(IMF) balance of payments statistics are the only ser-vices trade statistics available on a global basis, andcapture cross-border supply, consumption abroad (aspart of the category “travel”), and some temporarymovement of service suppliers. The more limited UnitedNations Conference on Trade and Development (UNC-TAD) data on FDI in services capture the flows throughwhich commercial presence is established. Finally, theUnited States is the only country that has regularly col-lected data on the sales of services by foreign affiliates.

4. The United States is the only country that hasregularly compiled data on sales of services to foreignpersons by majority-owned foreign affiliates of U.S.companies, and on sales of services to U.S. persons by

majority-owned U.S. affiliates of foreign companies. Acomparison of the balance of payments and foreign af-filiates transactions reveals in broad terms the relativeimportance of sales through cross-border delivery andcommercial presence.

5. It must be borne in mind, though, that the rela-tive importance of trade by different modes in a par-ticular sector reflects the choices of economic agentsgiven the constraints of both technological feasibilityand policy restrictions.

6. The FDI data are extremely thin, with data miss-ing for many countries and only available for three SSAcountries.

7. See the National Association of Software andService Companies (NASSCOM) Web site <http://www.nasscom.org>. These exports consist mainly of stan-dardized coding and testing services.

8. This report was prepared by McKinsey andCompany for NASSCOM.

9. These figures were computed from WTO 1998,table 3. Data refer to 1997.

10. See http://www.nasscom.org. The dominance ofon-shore delivery is due, among other things, to a re-duction in information asymmetries with regard to theperformance of programmers, the need for continuousclient-developer interaction, and demands by Indianprogrammers to be sent abroad, in part to improvetheir skills and expose themselves to international mar-kets (see Heeks 1998).

11. With permanent movement, the gains to thehost country must be weighed against the possible costto the home country in terms of “brain drain.” Over 50 percent of all migrating physicians come from de-veloping countries. In Ethiopia, for example, during1984–94, 55.6 percent of the pathology graduates fromthe Addis Ababa Faculty of Medicine left the country.In Ghana, of the 65 who graduated from the MedicalSchool in 1985, only 22 had remained in the country by1997. If these countries had adequate medical staff athome, these figures would be less cause for concern.

12. Other barriers to movement of natural personsinclude double taxation, wage-matching requirements(wages paid to foreign workers should be the similar tothose paid to nationals in that profession, eliminatingthe cost advantage for foreigners), and local trainingrequirements (to replace foreign with national laborwithin a certain time frame).

13. This is strictly true in static models withoutmarket imperfections—such as monopolistic marketstructures, internal and external economies of scale, orother distortions. The presence of imperfections opensup a plethora of possibilities in which the effects oftrade policies are typically indeterminate, dependingon the prior distortion.

14. See Hoekman and Braga (1997) for a review.


90

15. Consider, for instance, the case of the Arab Re-public of Egypt, where the import-weighted tariff was31 percent in 1997, and the average manufacturing-wide effective rate of protection (ERP) was much higherat 70 percent (Hoekman and Djankov 1997). However,services inputs used by Egyptian industry, including con-struction, communications, financial, business, distribu-tion, transport, and storage, were more expensive thanthey might have been if competition had been allowed.If it were assumed that prices were higher by, say, 15percent, then the average ERP for manufacturing wouldnot only be lower, but negative for several industries(chemicals, crude petroleum, and other extractive in-dustries), implying that the tariff on intermediate goods,together with the implicit tariffs on services inputs, out-weighed the tariff protection on the final goods.

16. If no trade in either goods or services is possi-ble, the production of final goods is cheaper in largermarkets, because a larger market can support a greatervariety of services. If trade in only goods is possible(for instance if services must be supplied through alocal establishment), then goods production tends toagglomerate in the larger country. The large countrygains from this as productivity increases since a largerfinal goods sector can support a wider variety of inter-mediate goods production. For the same reason, thesmaller country can lose from goods trade as final goodsproduction shrinks. However, if there is free cross-border trade in services, then all countries have accessto the full range of producer services. As a result, pro-ductivity in final goods production increases in allcountries, and so all countries gain from trade.

17. The last few years have seen a profusion ofnational and global computable general equilibriummodels seeking to estimate the economywide effects ofservices liberalization. The models suffer from weak-nesses, particularly the inadequate treatment of differ-ent modes of supply, the poor data on the levels of pro-tection in different services sectors, and an inability tocapture the regulatory institutional detail that is a keydeterminant of the consequences of services liberaliza-tion. The models are, nevertheless, useful in providinga rough idea of the costs of maintaining services barri-ers and the corresponding welfare gains from theirremoval.

18. As pointed out by Rodriguez and Rodrik (1999),there are two contradictory impulses on growth ema-nating from the scale effect described above. Protectinga sector increases its size, leading to higher growth, butit also creates a wedge between domestic and foreignprices imposing a production inefficiency that rises overtime exerting a negative impact on growth. The largerthe size of the protected sector the larger this impact. Bycontrast, liberalization of the services sector, in whichthe country is assumed to have a comparative disadvan-

tage, will also lead to increased static efficiency. Thiswill strengthen the growth impact of liberalization.

19. For example, there is evidence to suggest thatforeign bank entry qualitatively changed Turkish bank-ing by introducing financial and operations planningand improving the credit evaluation and marketing sys-tem (Denizer). Foreign banks also took the lead inspreading electronic banking and introduced new tech-nologies. They raised the human capital level of Turk-ish workers through domestic training programs, andby sending local recruits to training centers abroad.

20. Coe, Helpman, and Hoffmaister (1999) andLumenga-Neso and others (2001) are among thosewho present empirical evidence demonstrating the im-pact of technology diffusion—in their case throughtrade in goods—on total factor productivity growth. Inprinciple, the same should hold true for technologythat is diffused through factor flows

21. In terms of the financial instruments employed(too much reliance on short-term bills), in terms of thefinancial intermediaries that were unwittingly encour-aged (lightly regulated trust subsidiaries of the banks,and other newly established near-bank financial inter-mediaries), and in terms of market infrastructure de-velopment (failure to develop the institutions of thelong-term capital market). See, for instance, Claessensand Glaessner (1999)

22. Sometimes the object is to ensure access to allconsumers at the same price, irrespective of the cost ofprovision (for example, in transport and postal ser-vices). At other times, the object is ensure cheaper ac-cess for certain categories of users (for example, in fi-nancial services).

23. Mosely (1999) estimates the impact of financialliberalization on access to rural credit in four Africancountries Uganda, Kenya, Malawi, and Lesotho. Usingsample survey data, Mosely reports that between 1992and 1997, the percentage of sampled households withaccess to rural credit rose in Kenya and Uganda from13.1 percent and 9.2 percent to 25 percent and 21 per-cent respectively. However, in Malawi, there was in adecline in the corresponding number from 12 to 8 per-cent. Access to credit of the poorest 10 percent (by in-come) remained unchanged in Uganda and Kenya, butin the case of Malawi and Lesotho declined from 1.9 and 2 percent to .9 and 1.9 percent respectively.Mosely’s study also shows that financial reform by wayof financial innovation in rural areas and developmentof financial institutions catering to the poor has strongand significant effects on improving access to ruralcredit and lowering poverty. But simply privatizingstate micro-finance agencies has proven to be unsuc-cessful, as illustrated by the experience of Malawi.

24. The poor are likely to be unskilled, so the ques-tion arises as to the services sectors in which they are


91

likely to be employed. Unfortunately, data on the skillcomposition of the workforce in services sectors areonly available for some OECD countries and that at arather aggregate level. Still a certain pattern can be in-ferred. Construction, distribution and personal servicestend to be unskilled–labor intensive, whereas commu-nications, financial and business services tend to beskilled–labor intensive.

25. In India, the incumbent operator—the depart-ment of telecom expanded its workforce over the1996–2000 period. In the face of competition, it wasforced to improve its marketing strategy, expand itsnetwork and opened up thousands of public call officesall over India.

26. One such possibility is the case of “nonsustain-ability” of natural monopoly. This could arise, for in-stance, under some natural monopoly cost conditions,when there exist no prices that will not attract entry,even though single firm supply is efficient. Armstrongand others (1994, p. 106) conclude that, “Notwith-standing the logical possibility of this happening, we are doubtful whether it provides a good case forentry restrictions in the utility industries, which are not for the most part remotely contestable and wherethere is little evidence that cost conditions give rise tononsustainability.”

27. Interesting evidence in this context is availablefrom the Indian telecommunications sector. Das (2000)estimates a frontier multi-product cost function of theincumbent fixed-line operator, covering 25 years from1969 to 1994. The study finds the existence of veryhigh economies of both scale and scope in the technol-ogy used—the parameter estimates even suggest thattelecommunications in India is a natural monopoly.However, the incumbent operator displays great ineffi-ciency, leading to a 26 percent increase of the opera-tor’s cost of production. Based on these findings, Dasconcludes that India’s market liberalization program,started in the mid-1990s, is justified, but he argues thatthere may be a need to regulate entry in order to reduceunnecessary duplication of common costs. Moreover,with continued improvements in technology, the fixedcosts of entrants are likely to fall, reducing losses ofscale economies and thus increasing the costs of entryrestrictions.

28. Several countries have found it difficult to createan open, competitive telecommunications sector be-cause of a weak regulatory environment. Poland openedup its telecommunications sector to private competitionas early as 1990. There was a rush to invest, and about200 licenses were awarded in the first six years of thenewly liberalized regime. The dominant state operator,operating in a weak regulatory system, limited access toits network and benefited from unequal terms for rev-enue sharing, however. By 1996, only 12 of the 200 li-

censes were still being used by the few competitive op-erators who had managed to survive.

29. Studies of Argentina show that all incomeclasses gain from services reforms but that the rich (andthe foreign investors) gain relatively more if the regula-tor is weak and that the poor win relatively more if theregulator is effective in ensuring that the rents of thesector are shared with the rest of the economy (Chisariand Romero 1999; and FIEL 2000). The additionalgains from good regulation are estimated to be about0.35 percent of GDP on an annual basis.

30. As UNCTAD and World Bank (1994) argue, “Service providers are likely to prefer the higher in-comes that result from control of entry into their occu-pation, or form restrictions on competition betweenthose who are admitted to it . . . whenever regulation isjudged necessary, a major concern must be to ensurethat regulatory powers are not captured by the exist-ing providers of a service and used to further theirinterests.”

31. In some cases, though, where the cost of raisingrevenue is very high, the direct regulation route may bepreferable.

32. Hoekman 1996.33. For a detailed treatment see Mattoo 2000 and

forthcoming.

ReferencesAlexander, I., and A. Estache. 1999. “The Role of Reg-

ulatory Reform and Growth: Lessons from LatinAmerica.” Paper presented at TIPS AnnualForum. Johannesburg. September.

Armstrong, M., S. Cowan, and J. Vickers. 1994. Reg-ulatory Reform: Economic Analysis and BritishExperience. Cambridge, Mass.: MIT Press.

Barth, J. R., G. Caprio Jr. and R. Levine. 2000a. “TheRegulation and Supervision of Banks Around theWorld: A New Database,” Policy Research Work-ing Paper #2588, Development Research Group.Washington, D.C.: The World Bank.

———. 2000b. “Banking Systems Around the Globe:Do Regulation and Ownership Affect Perfor-mance and Stability?” Policy Research WorkingPaper #2325, Development Research Group. :The World Bank, Washington, D.C..

———. 2001. “Bank Regulations and Supervision:What Works Best?” Paper presented at the “13thAnnual World Bank Conference on DevelopmentEconomics.” April.

Bhagwati, J. N. 1984. “Splintering and Disembod-iement of Services and Developing Nations.” TheWorld Economy 7: 133–44.

Chadha, R., D. Brown, A. Deardorff, and R. Stern.2000. “Computational Analysis of the Impact on


92

India of the Uruguay Round and the ForthcomingWTO Trade Negotiations.” Discussion Paper No.459, School of Public Policy. The University ofMichigan, Ann Arbor.

Chisari, O., and C. Romero. 1999. “Winners andLosers from the Privatization and Regulation ofUtilities: Lessons from a General EquilibriumModel of Argentina.” The World Bank EconomicReview 13 (2).

Claessens, S., and T. Glaessner. 1999. “International-ization of Financial Services in Asia.” Paper pre-sented at the conference “Investment Liberaliza-tion and Financial Reform in the Asia-PacificRegion,” Sydney, Australia. August.

Coe, D. T., E. Helpman, and A. W. Hoffmaister. 1997.“North-South R&D Spillovers.” The EconomicJournal 107 (440): 134–49.

Copeland, B. R. 2001. “Benefits and Costs of Tradeand Investment Liberalization in Services: Impli-cations from Trade Theory.” Paper prepared forthe Department of Foreign Affairs and Interna-tional Trade, Government of Canada.

Cowhey, P., and M. M. Klimenko. 1999. “The WTOAgreement and Telecommunication Policy Re-forms.” Draft report for the World Bank. Gradu-ate School of International Relations and PacificStudies, University of California at San Diego.March.

Dages, G. B., L. Goldberg, and D. Kinney. 2000. “For-eign and Domestic Bank Participation in Emerg-ing Markets: Lessons from Mexico and Argen-tina.” Economic Policy Review 6(3):17–36.Federal Reserve Bank of New York. September.

Das, N. 2000. “Technology, Efficiency and Sustain-ability of Competition in the Indian Telecommu-nications Sector.” Information Economics andPolicy 12: 133–54.

DeFreitas, D., C. Kenny, and R. Schware. 2001.“Caribbean Cooperation: Rise of the RegionalRegulator.” Journal of Policy, Regulation andStrategy for Telecommunications Information andMedia 3: 195–201.

Denizer, Cevdet. 2000. “Foreign Entry in Turkey’sBanking Sector, 1980–1997.” Chapter 14 in S.Claessens and M. Jansen, editors, International-ization of Financial Services: Issues and Lessonsfor Developing Countries. Boston: Kluwer.

Eichengreen, B. Forthcoming. “Capital Account Liber-alization: What do the Cross-Country Studies TellUs?” World Bank Economic Review.

Estache, A., Q. Wodon, and V. Foster. 2001. “Ac-counting for Poverty in Infrastructure Reform:Learning from Latin America’s Experience.”World Bank, Washington, D.C.

Ethier, W. 1982. “National and International ReturnsTo Scale in the Modern Theory of InternationalTrade.” American Economic Review June 72:492–506.

Fink, C., A. Mattoo, and R. Rathindran. 2001. “Lib-eralizing Basic Telecommunications: The AsianExperience.” Paper presented at a workshop on“Trade, Investment and Competition Policies inthe Global Economy.” Hamburg Institute of In-ternational Economics. January.

FIEL (Fundacion de Investigaciones Economicas Lati-noamericanas). 2000. “La Regulacion de la Com-petencia y de los Servicios Publicos: Teoria y ex-periencia Argentina Reciente.” Buenos Aires.

Glassman, C. A. 2000. “Customer Benefits from Cur-rent Information Sharing by Financial ServicesCompanies.” A study for the Financial ServicesRoundtable. December.

Heeks, R. 1998. “The Uneven Profile of Indian Soft-ware Exports.” Working Paper No 3, Institute forDevelopment Policy and Management. Universityof Manchester, U.K.

Hodge, J. 1999. “Liberalizing Communications Ser-vices in South Africa.” Trade and Industrial Pol-icy Secretariat, Johannesburg.

Hoekman, B. 1996. “Assessing the General Agreementon Trade in Services.” In The Uruguay Roundand the Developing Countries, W. Martin, and L. A. Winters, editors. Cambridge, U.K.: Cam-bridge University Press.

Hoekman, B., and S. Djankov. 1997. “Effective Pro-tection and Investment Incentives in Egypt andJordan: Implications of Free Trade with Europe.”World Development 25: 281–91.

Hoekman, B., and C. Primo Braga. 1997. “Protectionand Trade in Services: A Survey.” Open Econom-ics Review 8: 285–308.

Kitchenman, W. F. 1999. “U.S. Credit Reporting: Per-ceived Benefits Outweigh Privacy Concerns.” TheTower Group.

Konan, D., and K. E. Maskus. 2000. “Service Liberaliza-tion in WTO 2000: A Computable General Equi-librium Model of Tunisia.” February. Processed.

La Porta, R., F. Lopez de Silanes, and A. Shleifer. 2000.“Government Ownership of Banks.” Harvard In-stitute of Economic Research Discussion PaperSeries (U.S.). 1890:1–55. March.

Lumenga-Neso, O., M. Olarreaga, and M. Schiff.2001. “On Indirect Trade-Related Research andDevelopment Spillovers.” Policy Research Work-ing Paper No. 2580. World Bank, Washington,D.C.

Markusen, J. R. 1989. “Trade in Producers Servicesand in Other Specialized Intermediate Inputs.”American Economic Review 79: 85–95.


93

Markusen, J., T. F. Rutherford, and D. Tarr. 2000.“Foreign Direct Investment in Services and theDomestic Market for Expertise.” Policy ResearchWorking Paper No. 2413, World Bank, Washing-ton, D.C.

Mattoo, A. 2000. “Developing Countries in the NewRound of GATS Negotiations: Towards a Pro-Active Role.” World Economy 23 (4): 471–489.

———. Forthcoming. “Shaping Future Rules for Tradein Services: Lessons from the GATS.” In T. Ito,and A. O. Krueger, Trade in Services. Cambridge,Mass.: NBER.

Mattoo, A., and L. Schuknecht. 2000. “Trade Policiesfor Electronic Commerce.” World Bank PolicyResearch Working Paper No. 2380. World Bank,Washington, D.C.

Mattoo, A., R. Rathindran, and A. Subramanian.2001. “Measuring Services Trade Liberalizationand its Impact on Economic Growth: An Illus-tration.” World Bank Policy Research WorkingPaper No. 2655. World Bank, Washington, D.C.

Mosely, P. 1999. “Micro-Macro Linkages in FinancialMarkets: The Impact of Financial Liberalizationon Access to Rural Credit in Four African Coun-tries.” Journal of International Development 11:367–384.

Petrazzini, B. A., and P. Lovelock. 1996. “Telecommuni-cations in the Region: Comparative Case Studies.”Paper presented at the “International Institute forCommunication Telecommunications Forum.”Sydney, Australia. April 22–23.

Robinson, S., Z. Wang and W. Martin. 1999. “Capturingthe Implications of Services Trade Liberalization.”Paper presented at the Second Annual Conferenceon Global Economic Analysis, GL Avernaes Con-ference Center, Ebberup, Denmark. June 20–22.

Rodríguez, F., and D. Rodrik. 1999. “Trade Policy andEconomic Growth: A Sceptic’s Guide to the

Cross-National Evidence.” Discussion Paper No.2143, Center for Economic Policy Research. May.

Rubin, Howard A. 1999. Global Software Economics.Hunter College, Department of Computer Sci-ence, and Rubin Systems Inc. New York.

Sampson, G., and R. Snape. 1985. “Identifying the Is-sues in Trade in Services.” World Economy 8(June): 175–82.

Smith, P. 1995. “End of the Line for the Local LoopMonopoly.” Public Policy for the Private Sector,Note No. 63. December. World Bank, Washing-ton, D.C.

UNCTAD (United Nations Conference on Trade andDevelopment) and World Bank. 1994. “Liberaliz-ing International Transactions in Services, AHandbook.” New York and Geneva.

UNCTAD (United Nations Conference on Trade andDevelopment) and WHO (World Health Organi-zation). 1998. International Trade in Health Ser-vice—A Development Perspective. SimonettaZarilli and Colette Kinnon, editors. Geneva:World Health Organization.

Wellenius, B. 1997. “Telecommunications Reform—How to Succeed.” Public Policy for the PrivateSector, Note No. 130. October. World Bank,Washington, D.C.

World Bank. 1995. Global Economic Prospects. Wash-ington, D.C.

———. 2001a. Can Africa Claim the 21st Century?Washington, D.C.

———2001b. Finance for Growth: Policy Choices in aVolatile World. New York: Oxford UniversityPress.

WTO (World Trade Organization). 1998. “GeneralAgreement on Trade in Services: Results of theNegotiations on Financial Services.” March.


94

trade in services: using openness to grow

Documents