UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT

POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES

STUDY SERIES No. 10

REGIONAL TRADE AGREEMENTS AND DEVELOPING COUNTRIES:THE CASE OF THE PACIFIC ISLANDS’ PROPOSED

FREE TRADE AGREEMENT

by

Robert Scollay

APEC Study Centreand

Economics Department, University of AucklandNew Zealand

UNITED NATIONS

New York and Geneva, 2001

ii

NOTE

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

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

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

ChiefTrade Analysis Branch

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

Palais des NationsCH-1211 Geneva

UNCTAD/ITCD/TAB/11

UNITED NATIONS PUBLICATION

Sales No. E.01.II.D.16

ISBN 92-1-112527-8

ISSN 1607-8291

© Copyright United Nations 2001All rights reserved

iii

ABSTRACT

Fourteen of the world’s smallest and most vulnerable economies – the Forum Island Countries(FICs) of the Pacific island region – are in the process of forming themselves into a free trade area(FTA). This paper begins by reviewing the characteristics of the FICs and their external trade. It isshown that despite their small size this is in many ways a very diverse group of economies. Standardanalysis of FTAs suggests that trade creation effects from a FIC FTA are likely to be small and thatthere may be a substantial risk of trade diversion. Loss of tariff revenue is a major concern, whichneeds to be addressed by restructuring of tax and tariff systems in some cases. Quantitative studieshave confirmed the small size of the trade creation effects and indicated the size of likely tariff losses,but were somewhat reassuring on the issue of trade diversion. Studies have also highlighted theimportance of continuing attention to most-favoured-nation tariff reductions in parallel with the formationof the FTAs in order to ensure that welfare effects are positive. A brief outline of the proposed FTAis provided. The proposed FTA should not be evaluated as a “stand-alone” exercise but as part of awider process of gradually integrating the FICs into the global economy. It must also be seen in thecontext of the FICs’ existing non-reciprocal free trade arrangements with Australia and New Zealandand the European Union, and the prospective future development of those relationships on a reciprocalbasis.

iv

ACKNOWLEDGEMENTS

This paper was written while the principal author was a visiting scholar at UNCTAD’s Divisionon International Trade in Goods and Services, and Commodities in Geneva. The support of theDivision and its Director, John Cuddy, is gratefully acknowledged. Grateful thanks are also due to BijitBora and Lucian Cernat for their support and advice and for comments, and to an anonymous refereefor helpful comments on an earlier draft of the paper. Any remaining errors are of course theresponsibility of the author.

v

CONTENTS

INTRODUCTION ...................................................................................................................1

I. ECONOMIC CHARACTERISTICS OF THE FICsAND THEIR TRADE..................................................................................................3

A. Economic size and income levels...............................................................................3B. Production structures................................................................................................5C. Intra-FIC trade ........................................................................................................7D. Trade with non-FICs................................................................................................8E. Balance of trade.......................................................................................................9F. Tariffs ....................................................................................................................10G. WTO membership .................................................................................................10H. Trade preferences ..................................................................................................12I. Regional trade initiatives .........................................................................................12

II. ISSUES IN DEBATES OVER RTAs –RELEVANCE TO THE FIC FTA.............................................................................14

A. The traditional Vinerian analysis ..............................................................................14B. Tariff revenue.........................................................................................................16C. Attraction of foreign direct investment .....................................................................17D. Increased competition............................................................................................17E. Other arguments.....................................................................................................17F. Conclusions ...........................................................................................................19

III. EVALUATION OF THE FIC FTA ...........................................................................20

A. Conclusions ...........................................................................................................22

IV. THE FIC FTA PROPOSAL.......................................................................................24

V. LINKS TO PREFERENTIAL ARRANGEMENTS WITHDEVELOPED COUNTRY PARTNERS..................................................................26

VI. CONCLUSION ..........................................................................................................29

REFERENCES ......................................................................................................................30

vi

List of figures

1. Population of Forum Island Countries, 1993 or later ........................................................42. Nominal gross domestic product (1) ................................................................................53. Nominal gross domestic product (2) ................................................................................64. Nominal gross domestic product per capita......................................................................75. Trade with other FICs.....................................................................................................86. Ratio of total imports to total exports .............................................................................107. Tariff collections as percentage of imports......................................................................118. Tariffs as percentage of total tax revenue........................................................................11

List of tables

1. Sectoral composition of GDP ..........................................................................................62. All countries: estimated changes in key economic variables.............................................203. Summary of results for FIC Free Trade scenarios ..........................................................21

Appendixes

1. Economic characteristics of the FICs: GDP, per capita GDP,and population..............................................................................................................34

2. Tariffs FIC....................................................................................................................343. Notes on CGE modelling and results reported in Scollay (1998).....................................354. FIC exports and imports by destination and source........................................................38

1

The proliferation of regional trading ar-rangements (RTAs) was a prominent feature ofthe international trading system in the last dec-ade of the twentieth century. The World Bank(2000) notes that of the 194 agreements noti-fied to the General Agreement on Tariffs andTrade (GATT) or the World Trade Organiza-tion (WTO) since the GATT’s inception, 87were notified in the years from 1990. Laird(1999) notes that 45 agreements were notifiedin the years from 1995 to 1998, with an esti-mated 62 further agreements which had not yetbeen notified to the WTO by mid-1998. A newsurvey by WTO (2000) counted a total of 172RTAs currently in force with a further 68 un-der negotiation, some of which are designed toreplace existing RTAs.

Not surprisingly, these developmentshave been accompanied by a lively debate onthe effects and implications of RTAs. Thereare several strands to this debate. One strandconcerns the benefits and costs of RTAs fortheir members, in comparison with other alter-natives open to them. A related issue is theeffect of RTAs on non-members. A further areaof debate related to both these issues is the setof questions regarding how RTAs can be de-signed to maximize benefits and minimizecosts. There is also a long-running debate overwhether the spread of RTAs threatens to un-dermine the multilateral trading system basedaround the WTO, and a closely related debateon whether the WTO’s disciplines and proce-

dures relating to RTAs are adequate to the situ-ation. The implications, negative or otherwise,of the tendency for RTA developments to cen-tre around major trading nations has been aparticular focus of debate.

Developing countries have not stoodaside from the trend towards RTAs. Develop-ing countries in all major regions of the globehave been and continue to be participants orpotential participants in RTAs, and many par-ticipate simultaneously in several such agree-ments. The issues raised in wider debates havenaturally also been applied to the questions re-lating to the place of RTAs in the trade strate-gies of developing countries, and the contribu-tion which participation in RTAs may make tothe development process.

Debate on this last point has a long his-tory and not surprisingly the focus of the de-bate has tended to shift in line with changingideas on the relationship between trade and de-velopment. In earlier years, when the importsubstitution paradigm heavily influenced think-ing on development issues, proposals for RTAsamong developing countries often reflected aninterest in exploring a regional as distinct froma national approach to import substitution,through the creation of larger protected mar-kets for the import-substituting industries toexploit. More recently, as outward-lookingtrade strategies have increasingly become thenorm, RTAs have tended more often to be

INTRODUCTION*

* On 28 June 2001, the trade ministers of the Pacific Islands Forum member countries announced that they hadendorsed a proposal for the establishment of a free trade area between the island country members of the Forum (theForum Island Countries, or FICs), to be known as the Pacific Island Countries Free Trade Agreement (PICTA). At thesame time they endorsed a proposal for a framework agreement, providing for the future strengthening at an appropriatepace of trade and economic cooperation between all Forum members (including Australia and New Zealand), to beknown as the Pacific Agreement on Closer Economic Relations (PACER). These agreements will be recommended forapproval and signature to the Forum leaders, who will be meeting in Nauru in August 2001.

2

evaluated for their contribution to the more ef-fective integration of their members into theinternational economy. Debate also continues,however, on the relation between trade and eco-nomic integration and the development of thedomestic economy, for example through im-proved infrastructure facilities.

This paper aims to make a modest con-tribution to debates surrounding participationby RTAs in developing countries by consider-ing the arguments commonly aired in these de-bates in the context of a specific proposal toform a free trade area (FTA) among the islandnations of the South Pacific, the so-called Fo-rum Island Countries (FICs), which made thepolitical decision in late 1999 to proceed withthe negotiation of an agreement on a FTA, re-ferred to in this paper as the FIC FTA. In theprocess it will be seen that the advantages anddisadvantages of a given RTA proposal needto be assessed in the light of the economic char-acteristics of the participating countries andtheir trade. It will also become clear that grasp-ing the full implications of a proposal such asFIC FTA requires an understanding of how theproposed arrangement may interact with themembers’ trade relations with their major de-

veloped country economic partners.

The first section of this paper highlightskey economic characteristics of the FICs andtheir major trading and economic relationships.This is followed by a discussion of the factorswhich are likely to be important in assessingthe benefits and costs of the proposed FIC FTA,leading to some preliminary conclusions aboutits likely economic effects. These conclusionsare then tested against some rudimentary quan-titative analysis of the implications of the FICFTA, leading to some further conclusions aboutFIC FTA’s likely economic effects and aboutthe way the agreement should be designed inorder to maximize benefits and minimize costs.The main features of the proposed FIC FTAare then outlined and briefly discussed in thelight of the conclusions from the precedinganalysis. A further section deals with the link-ages between the FIC FTA proposals and theexisting trade agreements between the FICs andtheir major developed country trading partners,together with the complications which mayarise for the ongoing management of these ex-isting arrangements in the context of the FICFTA proposal. This section is followed by abrief concluding section.

3

A central theme of this paper is that theeconomic characteristics of the members of anRTA and the characteristics of their existinginternational trading relations have an impor-tant bearing on which issues are likely to beimportant in assessing the role of the RTA intheir trade strategy, the likely benefits and costsof the RTA, and possibly also on the likely bal-ance between those benefits and costs. Thepaper begins therefore with a brief review ofthe economic characteristics of the FICs andtheir existing trading relations. The data pre-sented in this section of the paper were col-lected mainly from national statistics agenciesduring visits to the FICs in the early part of1998, and in each case are the latest data avail-able at that time. In some cases the data are notavailable in published form.

There are fourteen FICs in all: theMelanesian States of Papua New Guinea, theSolomon Islands and Vanuatu; the MicronesianStates of the Federated States of Micronesia,Kiribati, the Marshall Islands, Nauru and Palau;the Polynesian States of the Cook Islands, Niue,Samoa, Tonga and Tuvalu, and the hybridPolynesian/Melanesian State of Fiji, which alsohas a large Indian population. All fourteen ofthese States are members of the Pacific IslandsForum, which also includes as members Aus-tralia and New Zealand, which have strong tra-ditional ties with the Melanesian andPolynesian FICs (including Fiji) in particular.

The Pacific Islands Forum (formerly theSouth Pacific Forum) provides a vehicle forcooperation among the FICs themselves, andbetween the FICs and Australia and New Zea-land as the two developed countries of theSouth Pacific. There is a somewhat uneasy bal-ance, or tension, between these two roles. The

Forum Secretariat provides the FICs with tech-nical and administrative support. For the FICsthe Forum is both an expression of the socialand cultural linkages extending far back intotheir history, and a means of renewing, strength-ening and deepening those linkages, as well asbuilding a foundation for closer economic re-lationships.

A. Economic size and income levels

The fourteen FICs are all extremelysmall economies by international standards, butat the same time there are also enormous vari-ations among them in relative size. Figures 1to 3 illustrate these points, and the data onwhich these figures are based are shown asappendix 1.

The total population of the fourteenFICs is just over 6 million, of which 4.14 mil-lion and 0.75 million respectively are accountedfor by Papua New Guinea and Fiji. The popu-lation of individual FICs ranges from PapuaNew Guinea’s 4.14 million to an estimated2,300 in Niue. Three FICs (Niue, Nauru andTuvalu) have a population of less than 10,000,and a further two (the Cook Islands and Palau)have populations of between 10,000 and20,000. While the land area of most FICs isvery small, large expanses of ocean separatethe FICs from each other, and in many casesalso separate the constituent islands of the in-dividual FICs. One consequence of this oce-anic separation is that most FICs have very largeExclusive Economic Zones, and the marineresources within these zones are among theirmost valuable resources. On the other hand,isolation, small size and susceptibility to natu-ral disasters, as well as severe fluctuations in

I. ECONOMIC CHARACTERISTICS OF THE FICSAND THEIR TRADE

4

the world prices of their main exports, are allelements of the vulnerability of the FICs assmall island States.

In terms of economic size, available sta-tistics indicate a total combined gross domes-tic product (GDP) of the fourteen FICs in themid-1990s of US$ 8,452 million. To put thisin perspective, this is approximately 13.5 percent of the GDP of New Zealand, the smallerof the two developed country Forum membersand one of the smallest members of the Or-ganization for Economic Co-operation andDevelopment (OECD). Of this total FIC GDP,US$ 7,112 million, or just over 83 per cent,was accounted for by Papua New Guinea andFiji alone. Figure 2 shows clearly the disparityin economic size between these two economiesand the remaining 12 FICs. Figure 3, whichexcludes Papua New Guinea and Fiji, allows amore meaningful comparison of the economicsize of the remaining 12 economies.

The very small size of the combined FICmarket suggests that the potential for economic

gains based on economies of scale throughforming an RTA is clearly limited. Further-more, small economic size is likely to be asso-ciated with severe limitations on availabilityof administrative resources. A regional tradearrangement which requires complex negotia-tions and administrative arrangements wouldabsorb a disproportionate share of those re-sources. It is unlikely that such an arrange-ment would be sustainable for the smaller FICeconomies.

There is also considerable variation inincome levels among the 14 FICs, as figure 4indicates. The range in GDP per capita is fromUS$ 8,204 in Palau to US$ 651 in Kiribati. Thisrepresents a ratio of 13:1 between the highestand lowest average income levels within theFICs, comparable in fact to the correspondingratios within the European Union or the NorthAmerican Free Trade Area (NAFTA). It isnoteworthy that the two largest economies,Papua New Guinea and Fiji, with per capitaGDP of US$ 1,111 and US$ 2,250 respectively,are not close to either end of the range. Such

0

500000

1000000

1500000

2000000

2500000

3000000

3500000

4000000

4500000

PNG F iji S olomonIs .

S amoa Vanuatu F SM T onga Kiribati MarshallIs .

Cook Is . Palau Nauru T uvalu Niue

Figure 1. Population of Forum Island Countries, 1993 or later

5

variation in income per capita may point to asignificant difference in relative costs of labourand capital and in skill levels in the labour force,and might normally be taken as an indicationof the potential for mutually beneficial trade totake place.

B. Production structures

The summary data in table 1 on thestructure of production in a number of FICs sug-gest, however, that this conclusion should notbe accepted too readily. Production in theseeconomies is dominated by agriculture, forestryand fishing. Agriculture, forestry and fishingaccounts for an especially large share of GDPin the Solomon Islands (41 per cent), Samoa(40 per cent) and Tonga (37 per cent). In mostother FICs the share is between 15 per cent and27 per cent. In Papua New Guinea mining ac-counts for 27 per cent of GDP, in addition tothe 26 per cent accounted for by agriculture,forestry and fishing. The share in GDP of ag-riculture, forestry and fishing is unusually lowin the Marshall Islands (13 per cent) and Palau

(7 per cent). The service sector accounts forthe largest share of GDP in most FICs. Serv-ices, excluding construction and electricity, gasand water, account for over 70 per cent of GDPin Palau (81 per cent), 79 per cent in Kiribati,75 per cent in the Marshall Islands and 73 percent in the Cook Islands, and for between 50per cent and 70 per cent of GDP in Tuvalu (67per cent), Vanuatu (64 per cent), Fiji (54 percent) and Tonga (50 per cent). The servicessector share of GDP is relatively low in PapuaNew Guinea (33 per cent) and Samoa (34 percent).

Manufacturing, on the other hand, hasnot developed much in most FICs. The shareof manufacturing in GDP is highest in Samoa(18 per cent) and Fiji (15 per cent), but the fig-ure for Samoa is heavily skewed by a singlelarge enterprise which exists solely to supplythe Australian and New Zealand automotiveindustries, taking advantage of preferential ac-cess available under the South Pacific RegionalTrade and Economic Cooperation Agreement(SPARTECA). In all other FICs for which dataare available, manufacturing accounts for 5 per

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

5500

PNG F iji S olomonIs .

Vanuatu F SM S amoa T onga Palau MarshallIs .

Cook Is. Kiribati Nauru T uvalu Niue

US$ m

illion

Figure 2. Nominal gross domestic product (1)

6

cent of GDP or less, except for Papua NewGuinea (8 per cent). This suggests immedi-ately that the range of manufactured goods thatare likely to be traded between the FICs in anRTA is probably very narrow, with most of the

supply potential residing in a single FIC, Fiji.The potential for trade in the agricultural, for-estry and fisheries products which dominateFIC production structures tends to be inhibitedby transport and quarantine problems, as well

0

50

100

150

200

250

300

SolomonIs.

Vanuatu FSM Samoa Tonga Palau MarshallIs.

Cook Is. Kiribati Nauru Tuv alu Niue

US$ m

illion

Figure 3. Nominal gross domestic product (2)

Table 1. Sectoral composition of GDP (%)

Cook Marshall SolomonIslands Fiji Kiribati Islands Palau PNG Samoa Islands Tonga Tuvalu Vanuatu

Agriculture, forestry, fishing 21.0 19.4 17.4 13.3 6.8 26.5 39.9 41.3 36.8 23.9 22.7Mining, quarrying 0.0 3.3 0.0 0.3 1.2 27.2 0.0 0.1 0.7 0.9 0.0Manufacturing 2.6 14.8 0.9 2.2 0.8 8.2 17.9 4.0 3.9 4.0 5.2Electricity, gas, water 2.9 4.1 1.2 2.5 1.4 1.3 6.4 1.8 2.5 3.6 1.7Construction 1.2 4.5 1.8 6.5 9.0 3.9 1.9 6.9 6.0 5.6 6.5Wholesale/retail, restaurants, hotels 20.1 16.5 11.2 17.4 35.0 8.6 10.4 10.1 13.3 19.0 32.9Transport, storage, communications 10.4 12.6 11.3 6.8 14.9 5.2 2.7 6.5 8.6 6.2 7.5Finance, insurance, real estate, business services 10.9 14.1 5.7 14.6 8.4 0.9 - 4.7 10.2 11.8 13.9Community/social/ personal services 27.9 17.4 36.9 34.4 21.6 13.1 20.8 23.9 22.6 30.2 11.6Adjustments 3.0 -6.7 13.6 2.0 0.9 5.1 0.0 0.7 -4.6 -5.1 -2.0

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Source: National Statistics.

7

as by the fact that the FICs tend to produce simi-lar products in these sectors.

Trade in services may offer a promis-ing avenue for development of trade betweenthe FICs, and Forum leaders have in fact al-ready expressed interest in extending the FICFTA to cover services. Service sectors such astourism may well benefit from closer integra-tion among the FICs. At the same time, sensi-tivities relating to land ownership and ethnicdifferences mean that issues such as right ofestablishment and mobility of business personswill have to be handled with great care.

C. Intra-FIC trade

An impression that the potential fortrade among the FICs may tend to be quite lim-ited is reinforced by data presented in figure 5,which shows the share of each FIC’s merchan-dise trade which is accounted for by other FICs(imports, exports and total trade). It is clearthat trade between FICs (intra-FIC trade) ac-counts for only a very small share of the FICs’total trade.

Specifically, the only FICs for whichtrade with other FICs accounts for more than10 per cent of total trade are Tuvalu (21 percent), Kiribati (18 per cent), and Samoa (13per cent). For half of the FICs, trade with otherFICs accounts for less than 3 per cent of totaltrade. These are the Solomon Islands (2.8 percent), Fiji (1.6 per cent), Nauru (1.3 per cent),the Marshall Islands (0.8 per cent), Papua NewGuinea (0.6 per cent), the Federated States ofMicronesia (0.1 per cent) and Palau (0.1 percent). It is significant that the two largest FICs,Papua New Guinea and Fiji, are both includedin this group.

Four FICs source more than 10 per centof their imports from other FICs: Tuvalu (21per cent), Kiribati (20 per cent), Samoa (14 percent) and the Cook Islands (11 per cent). Onthe other hand there are five FICs, includingPapua New Guinea and Fiji, which obtain 1per cent or less of their imports from fellowFICs: the Marshall Islands, Papua New Guinea,Fiji, the Federated States of Micronesia andPalau. The Solomon Islands and Nauru eachobtain between 3 per cent and 4 per cent of

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

Palau Cook Is. Niue Nauru F iji F SM MarshallIs .

T onga Vanuatu PNG T uvalu S amoa S olomonIs .

Kiribati

US$

Figure 4. Nominal gross domestic product per capita

8

their imports from other FICs. Exports to otherFICs do not account for more than 5 per centof the exports of any FIC. There are six FICsfor which exports to other FICs account for lessthan 1 per cent of total exports: Papua NewGuinea, Nauru, the Marshall Islands, the Fed-erated States of Micronesia, Palau and the CookIslands. There is also a group of six FICs forwhich exports to other FICs account for be-tween 3 per cent and 5 per cent of total ex-ports: Vanuatu, Tonga, Kiribati, Fiji, Tuvalu andSamoa.

Taken together with the narrow produc-tion base in most FICs, this information clearlyindicates that intra-FIC trade is unlikely in theforeseeable future to account for more than asmall share of total FIC trade, even if it in-creases substantially as a result of the estab-lishment of a free trade arrangement among theFICs. In addition to trade barriers, the low levelof intra-FIC trade of course also reflects theexistence of other significant obstacles to thistrade, particularly the high costs of transporta-

tion among the FICs, related both to their smallsize and to the large distances separating themfrom each other.

D. Trade with non-FICs

Data on the relative importance to theFICs of different import sources and exportdestinations are presented in the graphs in ap-pendix 4. The data are for the latest year forwhich the data could be obtained in the case ofeach FIC. They show that FIC imports arehighly concentrated on long-standing tradi-tional sources. Australia and New Zealand to-gether account for over 90 per cent of the im-ports of Nauru and Niue respectively, and 78per cent of Cook Island imports. They alsoaccount for between half and two thirds of theimports of seven other FICs: Tonga (67 percent), Samoa (60 per cent), Vanuatu (55 percent), Fiji (54 per cent), Kiribati (54 per cent),the Solomon Islands (50 per cent) and PapuaNew Guinea (56 per cent). In Micronesia there

0%

5%

10%

15%

20%

25%

T uvalu Kiribati S amoa Cook Is. Niue Vanuatu T onga S olomonIs .

F iji Nauru MarshallIs .

PNG F SM

Perc

enta

ge o

f tot

al

Exports Imports Total

Figure 5. Trade with other FICs

9

is also a dominant supply source – the UnitedStates – which accounts for 51 per cent of im-ports into the Marshall Islands and 40 per centof the imports of the Federated States of Micro-nesia.

Exports are much more diversified.New Zealand dominates the exports of Niue(98 per cent), but beyond that Australia andNew Zealand account for more than 50 per centof exports only in the case of Samoa (57 percent), with their share of Cook Island exportsbeing slightly below 50 per cent at 47 per cent.The two countries also enjoy a moderate shareof the exports of Fiji (24 per cent) and PapuaNew Guinea (26 per cent). Beyond that thereare relatively small shares of the exports ofTonga (18 per cent), Kiribati (9 per cent) andVanuatu (5 per cent).

There is also considerable variationamong the FICs in the relative importance ofother export markets. Japan dominates theexports of the Federated States of Micronesia(78 per cent), Palau, and Tonga (50 per cent),and accounts also for a significant share of theexports of the Solomon Islands (36 per cent),Papua New Guinea (17 per cent), Vanuatu (16per cent) and the Cook Islands (15 per cent).The European Union accounts for large sharesof the exports of Kiribati (57 per cent), Fiji (26per cent), the Solomon Islands (25 per cent),Vanuatu (22 per cent), Samoa (21 per cent) andPapua New Guinea (15 per cent). The UnitedStates has a significant share of the exports ofthe Cook Islands (25 per cent), Tonga (18 percent), Fiji (13 per cent) and Kiribati (13 percent). Other significant markets are the Re-public of Korea for the Solomon Islands (15per cent of total exports) and Bangladesh forVanuatu (30 per cent of total exports).

The variable importance of Australiaand New Zealand as export markets for the FICsis interesting in the light of the common ac-cess to these markets which the FICs have en-joyed under SPARTECA. This variability willreflect a mix of various factors affecting theability of the individual FICs to competitively

supply products to the two markets, as well asthe relative attractiveness of other markets.This experience may be an indication that therewill be similar variability in the ability or in-clination of the FICs to exploit opportunitiesarising from establishment of free trade amongthemselves.

E. Balance of trade

All but three of the FICs (Papua NewGuinea, Solomon Islands and Nauru) have adeficit in merchandise trade. The ratio of mer-chandise imports to merchandise exports forindividual FICs is shown in figure 6. For thoseFICs with merchandise trade deficits the ratioof imports to exports ranges from 636:1 and40:1 respectively for Tuvalu (not shown in fig-ure 6, and excluding re-exports) and Niue to1.6:1 for Fiji. In addition to Niue, there arefive FICs for which the ratio is between 10:1and 5:1 (Cook Islands, Samoa, Kiribati, Tongaand Palau). The ratio is between 5:1 and 3:1for Vanuatu, the Federated States of Micronesiaand the Marshall Islands.

The balance on services trade should ofcourse be considered as part of the overall bal-ance of trade, but unfortunately this informa-tion is not available for all FICs. Six FICs (Fiji,Kiribati, Samoa, Palau, Tonga and Vanuatu)enjoy surpluses on their services trade. In noneof these cases is the surplus on services tradesufficient to offset the merchandise trade defi-cit. Fiji and Palau have the largest surpluses onservices trade relative to the size of their mer-chandise trade deficits. Papua New Guinea,the Solomon Islands, Tuvalu and the FederatedStates of Micronesia all have deficits in theirservices trade. The services trade deficits ofPapua New Guinea and the Solomon Islandsare large relative to their merchandise tradesurpluses, and in the latter case result in anoverall trade deficit. The services deficits ofTuvalu and the Federated States of Micronesiaare additional to substantial deficits in merchan-dise trade.

10

F. Tariffs

Figure 7 shows FIC tariff revenues as apercentage of imports. This provides an im-plicit measure of the average tariff actuallyapplied to imports in each FIC. This implicitaverage tariff rate is considerably lower thanmight be expected on the basis of publishedtariff schedules in most cases, reflecting sub-stantial undercollection of tariffs. The reasonsfor undercollection in a number of FICs havebeen detailed in recent tariff studies by the Fo-rum Secretariat.

Figure 8 shows tariff revenues as a per-centage of total tax revenues in the FICs. Thisclearly indicates the importance of tariff rev-enues in the tax base of most FIC economies,although as noted later in this report many FICsare moving to implement tax reforms whichwill result in a lower share of total tariff rev-enue being provided by tariffs in the future. Theimpact of a FIC free trade arrangement on tar-iff revenues depends on the proportion of totaltrade which will be affected and the distribu-tion of that trade across different tariff classes.

More complete information on FIC tariff ratesis provided in appendix 2.

G. WTO membership

Three FICs – Fiji, Papua New Guineaand the Solomon Islands – are members of theWTO, and a further three – Samoa, Tonga andVanuatu – have applied for membership. Thisis enough to ensure that any RTA involvingthese FICs will have to comply with relevantWTO obligations. In the case of an RTA in-volving only FICs it can be sufficient to meetonly the relatively undemanding requirementsof the “Enabling Clause”, which essentiallyprovides a dispensation from normal WTOrules for various trade arrangements involvingdeveloping countries, including RTAs. An RTAincluding developed country members, on theother hand, must satisfy the requirements ofArticle 24 of the GATT, which is now part ofthe WTO Agreement, together with the “un-derstanding” on its interpretation incorporatedinto the Final Act of the Uruguay Round. Ifthe RTA also covers services trade it must com-

0

5

10

15

20

25

30

35

40

45

Niue Cook Is. Samoa Kiribati Tonga Palau Vanuatu FSM MarshallIs.

Fiji SolomonIs.

PNG Nauru

Figure 6. Ratio of total imports to total exports

11

ply in addition with Article V of the GeneralAgreement on Trade in Services (GATS). Fo-rum economic ministers have clearly stated acommitment to implement policies consistentwith WTO principles wherever possible. This

may be interpreted to imply that a FIC FTAshould not necessarily rely on a dispensationunder the Enabling Clause but rather aim tomeet the higher standards required under Arti-cle XXIV.

0

5

10

15

20

25

30

35

Vanuatu Kiribati S amoa T onga PNG F iji S olomonIs .

T uvalu Cook Is. Niue MarshallIs .

Palau F SM Nauru

Figure 7. Tariff collections as percentage of imports

0

10

20

30

40

50

60

70

Kiribati Vanuatu Tuvalu MarshallIs.

Samoa Fiji Tonga SolomonIs.

FSM PNG Palau Cook Is.

Figure 8. Tariffs as percentage of total tax revenue

12

H. Trade preferences

The FICs have historically enjoyed non-reciprocal trade preferences from their majortrading partners under a wide range of agree-ments, the most important of which have been:

• The Generalized System of Preferences(GSP), which is made available unilaterally bydeveloped countries to all developing countries,although the product coverage varies betweendeveloped country markets;

• The Lomé Convention, which providedpreferential access to the European Union,along with a range of other benefits for all FICmembers of the ACP (African, Caribbean andPacific) group of States. Until recently thisgroup excluded the Federated States of Micro-nesia, the Marshall Islands, Palau, Nauru, Niueand the Cook Islands. With the signing in 2000of the Cotonou Agreement to replace the LoméConvention, these six FICs have also becomemembers of the ACP group;

• The compacts of Free Association (CFA)with the United States, which provides tradeaccess privileges, as well various importantentitlements to financial assistance, for the Fed-erated States of Micronesia, the Marshall Is-lands and Palau;

• SPARTECA, which provides duty-freeaccess to the Australian and New Zealand mar-kets provided that rules of origin requirementsare met;

• The Papua New Guinea Australia Tradeand Commercial Relations Agreement(PATCRA), which is applicable only to PapuaNew Guinea.

These preferences are being heavilyeroded as the preference-granting countries re-duce their trade barriers in pursuit of both mul-tilateral and unilateral trade liberalization ob-jectives. There is also increasing pressure forexisting non-reciprocal preferential trade agree-ments with developed country partners to be

replaced by reciprocal arrangements. The Eu-ropean Union has proposed that the non-recip-rocal arrangements applying under the formerLomé Convention should be replaced by recip-rocal arrangements embodied in a RegionalEconomic Partnership Agreement (REPA), al-though it has left open the possibility that an-other format could be adopted (Commissionof the European Communities, 1997). TheCotonou Agreement is designed as a transi-tional arrangement, allowing time for new per-manent arrangements to be negotiated. It pro-vides for negotiations on the new arrangementsto begin in 2002.

It is noteworthy that among the prefer-ential regional trade agreements in which FICsparticipate, only SPARTECA and the CotonouAgreement include all 14 FICs in its member-ship.

I. Regional trade initiatives

There are also a number of preferentialregional trading initiatives within the FICgrouping. The most developed of these is thetrade agreement of the Melanesian SpearheadGroup (MSG), which originally involved threemembers of the MSG – Papua New Guinea,Solomon Islands and Vanuatu – and has morerecently been extended to include Fiji. Themembers of the MSG trade agreement haveagreed to eliminate tariffs on trade betweenthemselves for an agreed list of products(Grynberg and Kabutaulaka 1995). Initially,only a very small number of products was in-cluded in this agreement, but subsequent ne-gotiations have progressively expanded therange of products covered, which neverthelessremains quite restricted. While anecdotal evi-dence indicates some success in expandingtrade within the MSG group, the level of thistrade remains low. The MSG trade agreementdoes have a political significance in the con-text of the FIC FTA, in that it envisages theassertion of a distinct Melanesian identity, asagainst the Forum-wide concept embodied inthe proposed FIC FTA.

13

Fiji also has long-standing non-recipro-cal trade agreements with Kiribati and Tuvalu,and a more recent reciprocal agreement withTonga, in addition to its arrangements with theother MSG members. Like the MSG tradeagreements, Fiji’s bilateral trade agreements areall “positive list” agreements, which means thatthey apply only to a specified list of products.

These lists are generally rather short, particu-larly in the case of the bilateral agreement withTonga, which covers only a small number ofproducts. Short lists may reflect a reluctanceto commit to meaningful liberalization, al-though it could also reflect a relatively smallportfolio of products of export interest.

14

The FICs have already accepted theneed to participate in the worldwide trend to-wards freer trade by progressively liberalizingtheir trade policies. The appropriate way toevaluate the FIC FTA proposal is therefore onthe basis of the contribution it can make to theadoption of more liberal trade regimes, and tothe capturing of the economic benefits that canbe expected to follow from reductions in tradebarriers.

Over the years a considerable numberof factors have been proposed as having a con-siderable bearing on whether RTAs are likelyto be beneficial for members, the extent of anynegative effects on non-members, and whetherRTAs tend to undermine or support the multi-lateral trading system. Reference to the eco-nomic characteristics of the FICs can help toidentify those arguments which are more rel-evant and less relevant in the case of the pro-posed FIC FTA.

A. The traditional Vinerian analysis

Viner (1950) provided the essential in-sight that programmes for the removal of tradebarriers that might unequivocally improve thewelfare of the implementing country if imple-mented on a unilateral or multilateral basis, willnot necessarily do so if implemented on a pref-erential basis. As is well known, this is be-cause RTAs and other preferential agreements,by discriminating in favour of their members,must inevitably discriminate against non-mem-bers.

The essential logic of trade liberaliza-tion is that countries benefit by importing goodsand services which they are relatively ineffi-

cient at producing themselves, in exchange forthe goods and services which they can producerelatively efficiently. When trade liberalizationis undertaken on a most-favoured-nation(MFN) basis, either unilaterally or multilater-ally, these efficiency gains are ensured becausethe increased imports will come from the mostefficient international source, and the increasedexports will be of products in which the liber-alizing country is internationally competitive.MFN liberalization is thus a first-best policy.

Preferential trade agreements do enablemembers to enjoy gains from trade. Membersof such agreements will increase imports ofgoods and services which their partners canproduce more efficiently, and increase their ownexports of goods which they themselves canproduce more efficiently than their partners.This increase in trade is the trade creation ef-fect of a free trade area, and confers economicbenefits in the same way as the increase in traderesulting from non-discriminatory trade liber-alization. The gains from trade in a preferen-tial trading arrangement will, however, gener-ally be less than the gains available from a cor-responding reduction of trade barriers on a non-discriminatory basis, since such arrangementsare unlikely to include the most competitivesuppliers of all goods and services in theirmembership. This is especially likely to be trueof free trade areas involving small groups ofcountries, particularly if those countries arethemselves small.

Second, the gains from trade creationin a preferential trade arrangement can be partlyor wholly negated by an effect called trade di-version, which does not arise in the case of non-discriminatory trade liberalization. Trade di-version occurs when the preferences created

II. ISSUES IN DEBATES OVER RTAS –RELEVANCE TO THE FIC FTA

15

under the arrangement cause imports to beswitched from non-partner to partner countries,even though they are available from the non-partner countries at a lower cost. This can hap-pen of course because the higher-cost goodsfrom the partner country enter free of duty andother restrictions under the free trade agree-ment, which may allow them to be sold to cus-tomers at a lower price than the lower-costimports from the non-partner, to which dutiesand other restrictions continue to be applied.Although consumers may benefit from lowerprices, the resource cost to the importing coun-try of acquiring the goods, as measured by theforeign exchange outlay needed to pay for them,has increased.

The potential for negative effects comesabout as a result of the violation of the non-discrimination principle, which inevitablymeans that regional trade liberalization is a sec-ond-best economic policy. Whether the eco-nomic effects of second-best policies are posi-tive or negative generally depends on the factsof each particular case, and this is indeed thecase with RTAs. The conventional analysis ofpreferential trading arrangements such as freetrade areas and customs unions indicates thatthey may either benefit or harm their members,depending on whether trade creation outweighstrade diversion or vice versa. In part the rela-tive size of trade creation and trade diversioneffects depends on the design of each agree-ment, but it can also be related to economiccharacteristics of the proposed members andthe structure of their international trade.

Laird (1999) reports the general viewthat the prospects for maximizing trade crea-tion and minimizing trade diversion will begreater the larger the shares of the members’ intheir partners’ pre-existing trade, the larger andmore diversified the partners’ economies, thecloser the partners’ domestic prices to worldprices, and the greater the initial non-uniform-ity of the partners’ tariff structures. It is quiteclear that none of these characteristics, with thepossible exception of the last, are found amongthe FICs.

Rather, the small size and very narrowproduction base of the FIC economies, togetherwith the low levels of existing intra-FIC tradeand the limited potential for increasing it, sug-gest that the potential for trade creation in aFIC RTA would be quite small. On the otherhand, the relatively high tariffs found amongthe FICs means that the margins of preferencecreated in favour of the members in a FIC RTAcould be very substantial, with the correspond-ing danger that a significant part of any in-creased trade under such an arrangement wouldconstitute trade diversion rather than trade crea-tion. Taken together, these factors suggest thatthe welfare effects of a FIC RTA are likely tobe very small, and there is a very real questionas to whether they are likely also to be nega-tive, owing to a preponderance of trade diver-sion over trade creation effects.

In any RTA the size of potential tradediversion effects can be reduced if the partnerscontinue to reduce their external trade barriers(i.e. trade barriers applying to imports fromnon-members) on an MFN basis at the sametime as they eliminate trade barriers betweeneach other. Where the apparent risk of tradediversion is high, as in the FIC case, this ac-companying MFN liberalization is an especiallyimportant condition for increasing the likeli-hood that the welfare effects of the RTA willbe positive. At the same as it ensures enhancedwelfare benefits for the members of the RTA,this MFN liberalization helps to limit the nega-tive impact on the exports of internationallycompetitive non-members, and thus also helpsto tilt the balance of global welfare effects to-wards the positive side.

MFN liberalization in isolation, ofcourse, is the first-best policy and would pro-duce unambiguously positive welfare effectsbecause of the absence of any potential for tradediversion. One can readily hypothesize that thesuperiority of this option in terms of overallwelfare effects is likely to be particularly pro-nounced in the case of the FICs. One couldalso hypothesize that an RTA which includessome of the FICs’ principal import sources,

16

such as Australia and New Zealand, would belikely to generate less trade diversion and wouldaccordingly yield superior welfare benefits tothe FICs compared with an RTA limited to theFICs themselves.

B. Tariff revenue

The heavy dependence of the FICs ontariffs as a revenue source for their Govern-ments is also quite clear, as is the fact that noGovernment can contemplate a sudden collapsein its revenue base. This is therefore a majorpotential constraint on trade liberalization bythe FICs.

One solution, of course, is to restruc-ture tax systems so as develop alternativesources of revenue and thus reduce the depend-ence on tariffs for revenue purposes. The FICsare aware of this and most are either moving orhave plans to move in this direction. Severalhave introduced a value-added tax as a meansof broadening their revenue base, and othersare contemplating the introduction of either avalue-added tax or some other broad-basedconsumption tax. In addition, improvementsin tariff administration and modernization ofcustoms procedures can be and are being madeto improve the revenue flow derived from anygiven tariff structure, for example by reducingthe extent of under-collection of tariffs.

Restructuring of tax and customs sys-tems is thus one of the major adjustments –perhaps the major adjustment – that the FICsare required to make to accommodate a liber-alizing trade strategy. The time needed for thisadjustment is likely to be a major limiting fac-tor for the speed with which the FICs are ableto undertake trade liberalization. Furthermore,the adjustment will be greater the more com-prehensive the liberalization, since this willimply a correspondingly large loss of tariff rev-enue. There is thus a clear inverse relation be-tween the size of the welfare gains to be ex-pected from any given liberalization and thesize of the adjustment that must be undertaken.

An initiative such as a FIC FTA would be ex-pected to produce small, possibly negativewelfare effects, but at the same time would havea relatively minor impact on tariff revenue col-lections. The more comprehensive liberaliza-tion involved in an RTA including a major trad-ing partner may be projected to yield muchlarger welfare gains, but it will also cause amuch larger loss of tariff revenue, and both ef-fects would be further accentuated in a moveto free trade on an MFN basis. The more com-plete the liberalization, the larger the adjust-ment, and the longer the period of time likelyto be needed to make that adjustment.

There is also the issue of the economicmeaning for countries such as the FICs of re-moving tariffs if the tariffs must be replacedby alternative taxes of equivalent revenue-gen-erating capacity. The advantage of a broad-based tax such as a value-added tax is that it isneutral as between imports and domestic pro-duction. In addition, it is relatively easy to zero-rate exports and thus remove the bias againstexports which derives from the effect of tariffson domestic cost structures, without any needfor recourse to complicated and administra-tively onerous drawback schemes. In suchcases the removal of tariffs still confers an eco-nomic benefit which follows from the removalof their distortionary effects.

This argument is less convincing underthe conditions that exist in some of the smallerFICs, where exports are negligible or even zero,and where almost all consumption is import-based. In these cases the replacement of thetariff by a value-added tax or similarly broad-based consumption tax may have little practi-cal significance except as a formal prerequi-site for participation in trade agreements whichrequire the removal of tariffs. The fact thatexports are negligible does not, however, meanthat they are not being inhibited by the pres-ence of tariffs. It is still possible that the re-moval of the anti-export bias inherent in thetariff could lead to the development of someexport potential.

17

C. Attraction of foreign direct investment

Ethier (1998) argues that the importanceof the traditional Vinerian analysis has beenovertaken by the new role of RTAs as the meansby which small countries seek to compete witheach other to attract foreign direct investment(FDI). It is certainly true that the desire to at-tract FDI has been an important motivation forthe formation of some recent RTAs among de-veloping countries, notably the Association ofSouth-East Asian Nations (ASEAN) and in-deed also for the participation of developingcountries in RTAs with developed countries,for example Mexico’s participation in NAFTA.It is certainly true also that the FICs, in com-mon with other developing countries, see anincrease in FDI as a hoped-for consequence oftheir economic reform programmes, of whichtrade liberalization forms an important part.

It is tempting to suggest that the forma-tion of a FIC FTA may lead to an increase inFDI by creating a larger potential market thanthat offered by individual FICs. This sugges-tion must be tempered, however, by recogni-tion that even in combination the FICs com-prise a very small market in international terms,and that the physical obstacles to trade betweenthe FICs will remain significant, even if all tradebarriers are removed, as noted earlier. It is thusnot realistic to expect that a FIC FTA will leadto a quantum leap in FDI into the FICs. Theeffect may be positive, but it is likely to be rela-tively small. The most noticeable impact maybe on the investment strategies of firms withinthe FICs themselves, rather than of larger mul-tinational operations that tend to be seekinglarger markets.

D. Increased competition

Reductions in X-inefficiency and elimi-nation of monopoly rents due to increased com-petition are traditionally identified as sourcesof economic gain in RTAs, and in some cir-cumstances these effects have been argued tobe quite large relative to other sources of ben-

efit. The tiny size of FIC markets means thatthe number of domestic suppliers to many mar-kets will be very small. It is easy to envisagethat a producer in, say, Tonga, faced with theprospect of greater competition from a coun-terpart in, say, Samoa, may be motivated to in-crease the efficiency of his or her operation.Thus to suggest that this may be a possiblesource of gain from the FIC FTA seems quitereasonable. Again, however, the gain is likelyto be very small, because the number of mar-kets for tradeable goods in which local produc-ers are significant suppliers is relatively small.

E. Other arguments

A consideration of the economic char-acteristics of the FICs and their trade leads quitereasonably to the suggestion that many of theother, sometimes very sophisticated ideas andarguments advanced to indicate possible eco-nomic benefits from the formation of RTAs willbe either irrelevant or at best of marginal im-portance to the analysis of a FIC FTA. Thus,for example, the tiny size of even the combinedFIC market is unlikely to provide significantopportunities for exploitation of economies ofscale, especially given the fact that a FIC FTAcannot affect the geographical isolation of eachindividual FIC. It is quite clear also that theFICs are not “natural trading partners” in thesense intended by Krugman (1991). They donot trade intensively with each other and theirgeographical proximity to each other is moreapparent than real as an economic factor; anypositive effect is likely to be nullified by thehigh transport costs for freight between small,widely separated islands. In any event, ofcourse, Bhagwati and Panagariya (1996) haveseverely criticized the notion that geographi-cal proximity provides a separate argument infavour of the creation of an RTA.

The “non-traditional arguments” iden-tified (and discounted) by Panagariya (1999),such as guaranteed access to markets and shel-ter from contingent protection, are likewise oflittle moment in the case of a FIC FTA. Any

18

markets to which guaranteed market access islikely to be of importance to the FICs lie intheir developed country economic partnersrather than in their fellow FICs, and likewisecontingent protection, to the extent that it is aproblem for the FICs at all, is likely to affectthem only in developed country markets. Theeconomic effects of a FIC FTA are likely to betoo small for it to have any significance in“locking in reforms”, another possible “non-traditional benefit” identified in Panagariya(1999). The exception to this comment is thatthe dependence of some FIC Governments ontariff revenue is so great that even the intro-duction of a FIC FTA may be sufficient to trig-ger far-reaching reform of tax systems in someFICs, and these reforms are likely to be irre-versible.

Lawrence (1997) suggests that the crea-tion of free trade areas tends to lead to encour-agement for deeper integration. Whether thisshould be counted as a benefit depends ofcourse on whether the integration arrangementis itself welfare-enhancing. Deeper integrationinvolving the harmonization of policies maynot be welfare-enhancing if the policy differ-ences being harmonized actually have a soundeconomic rationale. In any event, there are sig-nificant obstacles to most forms of deeper in-tegration among the FICs, although ambitionsin that direction clearly exist in the case of atleast some FICs.

On the other hand, the trade facilitationmeasures which increasingly accompany theliberalization packages in RTAs do typicallyyield enhanced welfare benefits. Scollay (1998)argues that if trade facilitation measures wereincluded in a FIC FTA, they could well be thelargest source of economic gain from the ar-rangement. The dominant share of agricultureand fisheries in the production structures ofmost FICs, together with the underdevelopedstate of quarantine services in the region, sug-gests that facilitation measures in the quaran-tine area might be particularly effective in en-couraging growth in beneficial trade.

Customs procedures are another tradefacilitation area where substantial improvementis possible. Although arrangements for coop-eration exist between the FICs in both the quar-antine and customs areas, Scollay (1998) iden-tifies a number of ways in which this coopera-tion could be substantially enhanced.

Liberalization of trade in services isanother aspect of deeper integration that mightwell be considered at a fairly early date as auseful extension of a FIC FTA, as noted ear-lier.

Given the tiny size of the FICs, a FICFTA will clearly have minimal economic ef-fects on non-members. Terms-of-trade effectswill be non-existent and although the risk oftrade diversion may be significant for the FICsthemselves, it will of negligible economic con-sequence for their trading partners. FIC FTAwill clearly not have anywhere near enougheconomic significance to strategically affect thebehaviour of other countries in the multilateraltrading system, although it may have some sig-nificance for the FICs’ own trade relations withtheir major developed country economic part-ners, as discussed later. It is sometimes arguedthat RTAs may impede unilateral or multilat-eral liberalization by creating vested intereststhat benefit from the preferences granted in theregional arrangement. However, given thelikely minor trade effects of a FIC FTA, anyvested interests created by its formation areunlikely to be large enough to seriously affecttrade policy in the FICs, let alone have any con-sequences for multilateral liberalization.

One possible negative effect of an RTAidentified in Panagariya (1999) is that mem-bers of an RTA may be tempted to raise MFNtariffs to compensate for revenue lost due tothe formation of the RTA. This policy is likelyto be at least partly self-defeating, since rais-ing MFN tariffs would tend to encourage fur-ther switching of imports to the RTA partners.In any event, the fact that the FICs already ap-pear to be moving towards restructuring theirtax systems to ensure sustainability of govern-

19

ment revenue in the face of trade liberalization,combined with their recognition of the impor-tance of their ongoing commitment to MFNtrade liberalization to their overall economicwelfare, means that it is also probably safe todiscount this possibility.

F. Conclusions

The conclusions from this section arequite clear. While there may be some minoreconomic benefits associated with increasedcompetition and encouragement of FDI, analy-sis of the benefits and costs of a FIC FTA re-duces in very large measure to the traditional

Vinerian concerns with the relative size of tradecreation and trade diversion effects, set againstthe constraints created by the FICs’ depend-ence on tariff revenue. Trade facilitation is theother area where substantial economic benefitsare possible, with services trade offering per-haps further opportunities in the future. Thereis also an important question as to how far anarrangement such as a FIC FTA serves as aneffective “stepping stone” to wider liberaliza-tion with greater potential for positive welfareeffects. However, suggestions that the Vineriananalysis has lost its relevance are clearly pre-mature, to say the least, in the case of econo-mies such as the FICs.

20

Scollay (1998) reports the results ofsome rudimentary computable general equilib-rium (CGE) simulations carried out to test someof the hypotheses identified in the previoussection regarding the economic effects of a FICFTA. A summary of the results of thesesimulations is given in table 2. Similarsimulations, using essentially the same data set,were performed by Stoeckel et al. (1998) foran FTA including Australia and New Zealandas well as the FICs; and, finally, Scollay andGilbert (1998) report the extension of the simu-lation exercise to consider full MFN liberali-zation by the FICs. For comparison purposes

the principal results from these three studiesare shown together in table 3.

Although widespread and serious datadeficiencies meant that construction of themodels used in these simulations was very time-consuming, they are nevertheless very crude.The crudeness of the models and the seriousdeficiencies in the data on which they are basedpreclude any serious weight being given to theparticular numbers generated in thesimulations. At best the results provide somepotentially useful information on the directionand order of magnitude of economic effects.

III. EVALUATION OF THE FIC FTA

Table 2. All countries: estimated changes in key economic variables

Percentage Percentage Percentage Percentage Percentage Change inchange in change in change change in change in welfare as a

imports from imports from in tariff total percentageFICs rest of world exports revenue employment of GDP

Cook Islands 7.88 -0.53 1.36 -11.24 0.26 -0.19Federated States of Micronesia 0.09 0.00 0.01 -0.01 0.00 0.00Fiji 7.72 0.22 0.14 -0.40 0.31 0.23Kiribati 13.32 -2.58 14.71 -21.87 0.80 -1.87Marshall Islands 10.02 0.07 0.49 -0.93 0.34 0.15Nauru 0.07 0.01 0.00 -3.55 0.01 0.02Niue 6.62 -0.13 8.81 -9.09 1.22 0.13Palau 3.43 0.00 0.01 -0.20 0.00 0.00Papua New Guinea 92.45 -0.34 0.04 -0.72 0.05 0.03Samoa 11.98 -1.04 4.73 -14.99 0.66 -0.68Solomon Islands 6.10 0.16 0.46 -3.41 0.43 0.20Tonga 10.54 -0.11 1.75 -7.40 0.59 0.01Tuvalu 3.14 0.27 6.94 -21.40 1.06 -0.01Vanuatu 2.67 2.58 5.06 -28.54 1.95 0.21

Additional simulations: reduction of rest-of-world tariff by 25%

Cook Islands 5.97 1.44 4.00 -31.63 1.90 0.60Kiribati 10.66 2.34 30.27 -37.00 3.82 0.45Samoa 9.76 2.92 11.49 -32.33 2.54 0.16Tuvalu 3.08 1.39 16.32 -38.61 3.27 1.16

21

A fuller description of the modelling processand its unavoidable limitations is given in ap-pendix 3.

One of the many difficulties was theshortage and in many cases almost completeabsence of reliable labour market data, or some-times even of any labour market data at all.Nevertheless, it seemed clear from examina-tion of the evidence available, including de-tailed interviews with responsible officials, thatconsiderable underemployment exists in theFICs, particularly in the subsistence sector. Inview of this, it seemed inappropriate to use thefull employment assumption typically used inCGE modelling, which implies that increasesin overall labour demand can be reflected onlyin wage increases. An assumption at the otherextreme would be that the level of unemploy-ment or underemployment is such that increasesin labour demand could be accommodatedthrough increases in employment, without anyneed for wages to rise. In the absence of anyreliable data on which an intermediate scenariocould be based, these two cases, which repre-sent opposite ends of the spectrum of possible

assumptions, were the two scenarios modelled.It could be hypothesized that reality might liecloser to the “unemployment” end of the spec-trum in the case of a FIC FTA, where the ef-fects on trade and production structures wouldbe relatively small, and progressively closer tothe “full employment end” of the spectrum asthe scope of the liberalization being modelledincreases. There is simply no way of beingmore precise than that.

Given the limitations of the results asoutlined above, it would be unwise to base anyconclusions too heavily upon them. They are,however, interesting and useful for the broadcorroboration they provide for a number ofhypotheses based on the standard analysis ofFTAs in the light of the economic characteris-tics of the FICs and their trade.

In the simulations reported in Scollay(1998) for the “unemployment” scenario, whichwas taken as the “base case”, the following arethe principal features of the results as shownin table 2:

Table 3. Summary of results for FIC Free Trade scenarios

All countries: estimated changes in economic welfare as a percentage of GDP

FIC free trade area With FIC free trade area Australia and New Zealand MFN liberalizationUnemployment Full Unemployment Full Unemployment Full

employment employment employment

Cook Islands -0.19 -0.30 2.61 -0.22 3.44 0.14Federated States of Micronesia 0.00 0.00 0.03 0.00 0.41 0.00Fiji 0.23 0.12 3.27 0.08 3.30 0.09Kiribati -1.87 -2.29 2.83 -1.75 7.15 1.80Marshall Islands 0.15 -0.07 1.05 -0.07 6.65 0.36Nauru 0.02 0.00 0.06 0.00 0.06 0.00Niue 0.13 -0.40 2.18 -0.39 4.59 0.21Palau 0.00 0.00 0.21 0.01 1.14 0.03Papua New Guinea 0.03 0.00 1.32 0.06 2.61 0.16Samoa -0.68 -0.84 1.06 -0.55 3.32 0.96Solomon Islands 0.20 -0.03 3.29 0.03 5.25 0.24Tonga 0.01 -0.25 1.73 -0.15 3.24 0.39Tuvalu -0.02 -0.54 - - 5.09 0.19Vanuatu 0.21 -0.42 0.90 -0.31 4.22 1.07

22

• All FICs increase imports from other FICs,which may be some indication of tradecreation (note that the high percentage in-creases in some cases can give a mislead-ing impression, since these increases aregenerally from a very low base).

• Imports from the rest of the world also in-crease for half the FICs, and the percent-ages of both increases and decreases arevery small, which suggests that trade di-version may not be as great a problem asmight have been feared.

• Exports rise is all cases (except one, wherethey remain constant).

• Employment rises in all cases.

• As expected, there are tariff revenue lossesin all cases, and in five cases the loss ismoderately severe.

• All but four FICs experience increases inwelfare, but the changes in welfare are gen-erally small (in all but two cases amount-ing to 0.2 per cent of GDP or less).

Scollay (1998) provides some commenton whether the projected increases in trade be-tween FICs are realistic. Although it is impos-sible to validate the results of the simulationsin full, a number of examples of potential in-creased trade are identified.

In the cases of the four FICs experienc-ing welfare losses under the initial simulations,as an experiment a further simulation was runin which external (MFN) tariffs were reducedby 25 per cent simultaneously with the forma-tion of the FTA. The results of these simulationsare also shown in table 2. In each case, this 25per cent MFN tariff reduction was sufficient toconvert the welfare losses into welfare gains,and furthermore, in each case also reductionsin imports from the rest of the world are con-verted into increases. This provided strikingconfirmation of the importance of ingoing MFNliberalization in parallel with any move to form

an FTA.

The comparison of the three differentcases in table 3 is also interesting for the con-firmation it provides of earlier hypotheses.While the “full employment” scenarios consist-ently produce less favourable results (as mightbe expected), a clear hierarchy emerges underboth scenarios. For almost every FIC the wel-fare gains increase (and in most cases increasesignificantly) as the coverage of the tariff re-moval widens from a FIC-only FTA to an FTAincluding Australia and New Zealand, and fi-nally to full MFN liberalization.

A. Conclusions

The analysis reported so far allows somefairly clear conclusions to be drawn about theproposal for a FIC-only FTA:

• The economic effects are likely to be verysmall, and may be negative for some FICs.

• It is critically important that MFN liber-alization be continued in parallel with theestablishment of the FTA. This will avertthe possibility of negative welfare effectsand enhance the size of likely positive wel-fare effects.

• The economic benefits are not likely to besufficient to justify a FIC-only FTA as anultimate objective of economic policy.Greater welfare gains are potentially attain-able through more comprehensive liberali-zation.

• Some FICs will experience significantlosses of tariff revenue even under a FIC-only FTA. The losses will be much greaterunder an FTA including Australia and NewZealand, and greater still under full MFNliberalization by the FICs.

• A FIC-only FTA may therefore have valueas an initial step in a liberalization strat-egy, requiring relatively modest adjust-

23

ments, and allowing time for the more far-reaching adjustments to be put in placewhich will be needed to accommodatemore comprehensive liberalization initia-tives.

• A FIC FTA should be supported by a strongprogramme of trade and investment facili-tation measures, especially in the quaran-tine and customs areas. The gains from thismay be as significant as, or more signifi-cant than, the gains from removing tariffsbetween members.

A further conclusion is that if a deci-sion is made to proceed with a FIC FTA, itshould be designed so as to minimize negotiat-

ing and administration costs, otherwise thesecosts will be very likely to outweigh the rela-tively minor economic benefits likely to comefrom the arrangement. Scollay (1998) stronglyrecommends a “negative list” approachwhereby trade is liberalized in all products ex-cept for a (preferably short) list of exclusions.This is contrasted with the “positive list” or“product-by-product” approach often used inthe negotiation of FTAs between developingcountries, which tends to be very intensive inthe use of negotiating and administrative re-sources. Under the “positive list” approach,furthermore, experience shows that progress inremoving trade barriers often slows down quiterapidly as the negotiations widen to take in in-creasingly sensitive products.

24

The decision by FIC leaders to proceedwith negotiation of a FIC FTA indicates thatthey accepted the argument outlined in Laird(1999) that the smaller immediate gains fromsuch an arrangement are an acceptable price topay to keep initial adjustment costs within tol-erable limits. In particular, the FICs appear tohave endorsed a “stepping stone” approachwhereby a small initial step allows collectiveliberalization to move forward while allowingtime to prepare for the adjustments likely toaccompany subsequent larger steps. The es-tablishment of a FIC FTA as an initial step alsoserves a “confidence building” and “capacitybuilding” function, providing an initial low-costtaste of the benefits of liberalization, and ex-perience in the negotiation and management oftrade agreements, thus helping to develop con-fidence and capacity to engage in further jointliberalization initiatives. Political considera-tions were also clearly important, notably theperceived need to provide a greater sense ofsolidarity among the FICs in international tradepolicy matters. This was seen as an importantcounterweight both to the variable participa-tion of FICs in most of their trade agreementswith developed countries, and to the tendencyto form trade coalitions on ethnocentric linesas reflected in the emergence of the MelanesianSpearhead Group trade agreement.

A draft negotiating text for a FIC FreeTrade Agreement has been prepared (Myburgh,1999). Its principal features are:

• Tariffs are to be phased out over 10 yearson an automatic timetabled basis.

• The approach to non-tariff barriers is thatthey are to be eliminated as and when theyare identified, if necessary through

tariffication. This approach was consid-ered preferable to spending resources ondeveloping and negotiating detailed rulesto address a problem which is consideredto be minor.

• A “negative list” approach is adopted, withprovisions to limit the size of negative listsand gradually phase them out.

• Rules of origin are not product-specific andprovide exporters with the option of usingthe “change of customs heading” (CCH)basis or a 40 per cent area content rule.

• “Emergency actions” (anti-dumping andcountervailing measures, safeguard andbalance-of-payment measures) are allowedunder strict conditions.

• An “infant industry” provision permits sus-pension of the agreement for nominatedproducts to allow “infant industries” to de-velop, again under strict conditions, includ-ing an automatic lapsing of the suspensionif the development does not proceed withina stated time frame.

• Principles for conduct of government pro-curement are included, although their im-plementation is non-binding and on a “bestefforts” basis.

• There is a three-stage dispute settlementprocess, progressing from consultation tomediation to arbitration.

• The operation of the agreement is to bereviewed after five years, and at regular in-tervals thereafter.

IV. THE FIC FTA PROPOSAL

25

The key considerations in these designfeatures are, first, to minimize negotiating andadministrative costs so that these offset anypositive welfare effects as little as possible, asexplained above. This is reflected in the auto-maticity built in to the tariff reduction sched-ules and especially in the adoption of a “nega-tive list” approach. Second, the provisions onemergency actions and infant industries reflectthe vulnerability of the FICs and the need toprovide for the possible future evolution of theirdevelopment strategies. The provisions are in-tended to ensure that the FICs have sufficientflexibility to deal with unusually severe eco-nomic disturbances and to implement freshdevelopment initiatives that may be proposedin future, while at the same time providing suf-ficient disciplines to ensure that this flexibilityis not used to undermine the agreement by re-introducing unjustified protection on a discre-tionary basis. Third, it is intended that theagreement will as far as possible meet the stand-ards laid down in GATT Article XXIV. Fourth,and relatedly, the agreement is designed as agenuine trade liberalization initiative, and thisis reflected in its wide coverage, the relativelyliberal rules of origin, and the automaticity ofits provisions.

The economic arguments in favour ofthis last point are not necessarily clear-cut.Laird (1999) points out that in the case of a

trade-diverting agreement it might be prefer-able to keep its coverage as narrow as possiblein order to minimize the economic damage –and there are some grounds to be concernedabout possible trade diversion in a FIC FTA,as noted earlier. However, since a FIC FTA isregarded as no more than a step in what is in-tended to be a more comprehensive liberaliza-tion process, it was felt that the arguments infavour of wide coverage should prevail. It wasalso considered that this could not be regardedas a threat to the development process in theFICs, in view of their very narrow productionbase, which means that there are relatively fewindustries or potential industries which will beput under pressure by the dismantling of pro-tection. These relatively few cases can be read-ily accommodated under the “negative list” pro-visions.

Following completion of the negotia-tions for establishment of a FIC FTA, it is en-visaged that detailed proposals will be devel-oped and implemented for trade facilitationmeasures to accompany the FTA, particularlyin the quarantine and customs areas. Exten-sion of the agreement to cover trade in serv-ices has already been mooted, and the FICs haveindicated a firm interest in extending the agree-ment to include the French and United StatesPacific territories.1 Studies on both the latterissues are being commissioned.

1 New Caledonia, French Polynesia, and Wallis andFutuna; Guam, Commonwealth of the Northern MarianaIslands, American Samoa.

26

The World Bank (2000) has suggestedthat developing countries may be better servedby entering into RTAs with developed coun-tries rather than with other developing coun-tries. However, the proposed FIC FTA is notan isolated venture into preferential trade onthe part of the FICs. The FICs already havelargely unrestricted and non-reciprocal duty-free access to both the Australian and NewZealand markets under SPARTECA and to theEuropean Union market under the Lomé Con-vention. The Federated States of Micronesia,the Marshall Islands and Palau (the “compactcountries”) have similar access to the UnitedStates market under the CFA. Furthermore, theFICs are facing what may turn out to be irre-sistible pressures to convert existing non-recip-rocal preferential arrangements with developedcountries into reciprocal arrangements.

The European Union, as is well known,and as noted earlier in this paper, has proposedthat the non-reciprocal preferential access ar-rangements extended over many years throughthe trade provisions of the Lomé Conventionshould be replaced by reciprocal market accessprovisions to be embodied in a series of Re-gional Economic Partnership Agreements(REPAs), which it plans to sign with groups ofcountries in the ACP regions. The CotonouAgreement, signed in mid-2000, replaces theLomé Convention and provides for negotiationson REPAs, or alternatives which might be pro-posed by the ACP States and accepted by theEuropean Union, to begin in 2002, with imple-mentation to begin in 2008. Australia and NewZealand have clearly indicated that as full mem-bers of the Pacific Islands Forum they expectto be included in the FIC FTA as full mem-bers, and the Forum leaders have in fact agreedthat a formula must be found which provides

appropriate “application of the [FIC FTA]measures to Australia and New Zealand”; theseneed not imply full membership, however. In-clusion of Australia and New Zealand in anFTA with the FICs may trigger provisions inthe Lomé Convention requiring that the FICsdo not grant any other developed countrygreater market access than they grant to the Eu-ropean Union, as well as similar provision inthe CFA with the United States for the threeformer United States trust territories. If the FICsgrant preferential access to the European Un-ion or the United States there is no formal le-gal requirement to grant equivalent access toAustralia and New Zealand, but the politics ofthe situation are such, given the latter two coun-tries’ position as full members of (and also themajor donors to) the Pacific Islands Forum, thatthe granting of such access would almost cer-tainly be unavoidable.

The design of new agreements to ac-commodate all these considerations, while atthe same time meeting the non-discriminationprovisions of the GATT and the relevant GATT/WTO rules on RTAs, raises many difficult is-sues. Some of the difficulties would be less-ened if a FIC RTA was established as a cus-toms union, with a common external tariff, thusmaking it easier for the FICs to enter new tradeagreements as a group. A customs union is,however, unlikely in the foreseeable future. Theissues involved in the design of the new agree-ments are beyond the scope of this paper, andwill not be discussed further here.

The proposed FIC FTA will thus be justone element in an interconnected web of pref-erential trade arrangements in which the FICsare involved. The web is developing in wayswhich will obviate any disadvantage the FICs

V. LINKS TO PREFERENTIAL ARRANGEMENTS WITHDEVELOPED COUNTRY PARTNERS

27

might have suffered from not opening up theirmarkets to developed country partners, andwhich in the process will of course also serveto further neutralize any further trade-divert-ing effects of a FIC FTA. The FIC FTA pro-posal also raises questions about how it will fitinto the overall architecture of trading arrange-ments in which the FICs are involved.

A FIC-only FTA causes relatively fewcomplications. It could be notified under the“Enabling Clause” to the Trade and Develop-ment Committee of the WTO, where it mightexpect an easy passage. The FICs would beleft free to negotiate new agreements with theirmajor developed country partners independ-ently of the provisions of the FIC FTA, althoughthey may wish to suggest that some provisionsof the FIC FTA be used as a model for theseagreements – other things being equal, it isclearly in the FICs’ interests to standardize asfar as possible the provisions of the preferen-tial agreements in which they are involved (ex-cept where differences are required in order totake account of development considerations),so as to minimize the administrative complexi-ties involved.

An FTA which includes Australia andNew Zealand, on the other hand will have tonotified under Article XXIV and examined bythe Committee on Regional Trade Agreements,where it will be subject to more rigorous scru-tiny, and where a number of potential difficul-ties in terms of Article XXIV may be raised.The market access provisions of the LoméConvention and the CFA would in this caseinevitably ensure that the negotiation and de-velopment of the FTA would become linked tothe FICs’ negotiations with the European Un-ion and the United States. If Australia and NewZealand are not included as full members ofthe FTA, an alternative formula for linking themto the provisions of the FTA will have to befound, both to satisfy the mandate from Forumleaders and to accommodate the political real-ity that Australia and New Zealand are boundto demand treatment equivalent to any marketaccess granted to the European Union in forth-

coming negotiations.

Regardless of the architecture adopted,it seems likely, in view of the legal claims and/or likely politically irresistible demands of themajor developed country economic partners forequivalent treatment in terms of market access,that the FICs will have to contemplate simul-taneously opening up their markets to Australiaand New Zealand, the European Union and theUnited States – and possibly also Japan, giventhat that country, a major aid donor in the re-gion, is unlikely to take kindly to discrimina-tion against it in favour of other developedcountries. In other words, in terms of the “step-ping stone” approach to liberalization, the stepafter a FIC FTA is likely to have to be a verylarge one, moving the FICs a considerable partof the way towards full MFN liberalization.

The previous discussion suggests that alarge second “step” of this nature would havea positive effect on the welfare gains from pref-erential liberalization, since it would help tocounter the trade diversion that might arise notonly in an FTA among the FICs but also if theFICs granted preferential access to only one oftheir developed country trading partners.Panagariya (1999) provides some grounds forapprehension about the latter possibility byoutlining a set of conditions under which asmall high-tariff country might suffer particu-larly large economic losses in an RTA with alarger low-tariff country. This could occur ifthe larger country failed to fully displace im-ports from other sources in the smaller coun-try’s market, and if under these conditions theprice of imports did not fall. In this case, theexporters in the larger country “capture” thetariff revenue previously received by thesmaller country’s Government. The disparatesize of the partners means that this loss willnot be offset by a similar effect operating ontrade in the other direction. The potential dan-gers of a situation such as this are avoided ifpreferential access is simultaneously grantedto each of the main sources of imports.

28

However, while a larger second “step”may be positive for the welfare gains from tradeliberalization, it also implies a correspondinglylarger adjustment, particularly in terms of theneed to provide for alternative sources of rev-enue to replace the lost tariff revenue. This islikely to affect the length of the time framewithin which the FICs are prepared to contem-plate taking this second step.

The FICs clearly face a major challengein managing interlinked negotiations with theirmajor developed country economic partners insuch a way that they retain control over the paceof their own liberalization, ensure that the re-sulting agreements do not involve them in un-necessary complexities or administrative bur-dens, and at the same time secure an accept-able level of compensation for the additionalmarket access they will be compelled to offer.

29

If the proposed FIC FTA is taken out ofcontext it can easily be presented as an initia-tive which makes little economic sense. Taken,however, in the context of the economic char-acteristics of the FICs and the trade policy is-sues which they face, it can be viewed as a smallbut useful element in an ongoing strategy of

trade liberalization. An understanding of theimplications of a FIC FTA, furthermore, re-quires an assessment of the ways in which itmay be linked to the trade negotiations whichthe FICs will be obliged to conduct with theirmajor developed country economic partners.

VI. CONCLUSION

30

Bhagwati, J. (1993), “Regionalism and Multilateralism”, World Economy 15, 535–555.

Bhagwati, J., D. Greenaway and A. Panagariya (1998), “Trading Preferentially: Theory andPolicy”, Economic Journal 108, 1128–1148.

Bhagwati J. and A. Panagariya (1996), ”Preferential Trading Areas and Multilateralism: Stran-gers, Friends or Foes?”, in J. de Melo and A. Panagariya (eds.), New Dimensions in RegionalIntegration (Cambridge, Cambridge University Press).

Commission of the European Communities (1997), Guidelines for the Negotiation of New Coop-eration Agreements with the African, Caribbean and Pacific (ACP) Countries (Brussels).

Ethier, W. (1998), “The New Regionalism”, Economic Journal 108, 1149–1161.

Grynberg, R. and T.T. Kabutaulaka (1995), “The Political Economy of Melanesian Trade Integra-tion”, Pacific Economic Bulletin, 48–60.

Kemp, M.C. and H. Wan (1976), “An Elementary Proposition Concerning the Formation ofCustoms Unions”, Journal of International Economics 6, 95–98

Krugman, P. (1991), “The Move to Free Trade Zones”, Federal Reserve Bank of Kansas CityEconomic Review, November/December, 5–25.

Laird, S. (1999), “Regional Trade Agreements: Dangerous Liaisons?, World Economy 22, 1179–1200

Lawrence, R. (1997), Regionalism, Multilateralism and Deeper Integration (Washington, D.C.,Brookings Institution).

McMillan, J. 1993, “Does Regional Integration Foster Open Trade? Economic Theory andGATT’s Article XXIV”, in K. Anderson and R. Blackhurst (eds.), Regional Integration andthe Global Trading System (London, Harvester Wheatsheaf).

Myburgh, P. (1999), Pacific Regional Trade Agreement (PARTA) and Protocols: Draft Negotiat-ing Text 1.0 (Suva, South Pacific Forum Secretariat).

Panagariya, A. (1998), “Do Transport Costs Justify Regional Preferential Trade Arrangements?No”, Weltwirtschaftliches Archiv 134, 280–301.