Report No. 24417-EAP Pacific Islands Regional Economic Report Embarking on a Global Voyage: Trade Liberalization and Complementary Reforms in the Pacific September 10, 2002 Poverty Reduction and Economic Management Unit East Asia and Pacific Region Document of the World Bank ACRONYMS AND ABBREVIATIONS ADB Asian Development Bank ACP Africa, Caribbean and Pacific CARICOM Caribbean Community EU European Union FDI Foreign Direct Investment FSM Federated States of Micronesia FTA Free Trade Agreement GNI Gross National Income GST Goods and Services Tax IMF International Monetary Fund MSG Melanesian Spearhead Group PACER Pacific Agreement on Closer Economic Relations PFTAC Pacific Financial Technical Assistance Centre PICs Pacific Island Countries PICTA Pacific Island Countries Trade Agreement REPA Regional Economic Partnership Agreement RTA Regional Trade Agreement SOE State Owned Enterprise SPARTECA South Pacific Regional Trade and Economic Cooperation Agreement VAGST Value Added Goods and Services Tax VAT Value Added Tax WTO World Trade Organization Vice President: Jemal-ud-din Kassum, EAPVP Country Director: Klaus Rohland, EACNF Sector Director: Homi Kharas, EASPR Task Manager: Vivek Suri, EACNF Table of Contents Executive Summary ....................................... .........................i 1. INTEGRATING WITH THE GLOBAL ECONOMY: TRADE STRATEGIES FOR THE PACIFIC ................................................1 1. A New Trade Regime for the Pacific ...............................................................l 1 1.1 Rationale for PICTA ................................................................1 2. Analyzing the Impacts of PICTA ................................................................5 2.1 Non-Economic Impacts of PICTA ................................................................ 5 2.2 Economic Impacts of PICTA ................................................................6 3. Quantitative Estimates of the Economic Impact of PICTA . . 16 3.1 Quantitative Estimates of Including Australia and New Zealand in the FTA ...... 17 4. Features of PICTA ............................................................... 20 5. Moving beyond PICTA ............................................................... 23 5.1 PICs and the EU ............................................................... 24 5.2 PICs, Australia and New Zealand ............................................................... 26 6. The WTO Alternative ............................................................... 30 7. Conclusion ............................................................... 31 2. INTERACTION OF TRADE LIBERALIZATION AND TAx REFORM .. 34 1. Introduction ............................................................... 34 2. The Structure of Revenue in Selected Pacific Island Countries . . 35 3. Impact of PICTA on Government Revenue - Estimates from Earlier Studies ... 36 3.1 The Timing of Revenue Impacts ............................................................... 40 4. Experience with a VAT in the Region ........................................... 42 4.1 Trends in Tariff and VAT revenue ............................................................... 42 4.2 Issues in VAT implementation ............................................................... 45 5. Reform Considerations in the Context of Trade Liberalization . . 47 6. Kiribati and Tonga: Potential Impact of PICTA and Policy Response . . 50 6.1 The Potential impact of PICTA ............................................................... 51 6.2 Current Tariff and Tax Systems in Kiribati and Tonga ........................................ 53 6.3 Administration and Capacity Constraints ............................................................. 55 6.4 Considerations for the Introduction of a Broad Based Consumption Tax ........... 56 7. Key Findings and Lessons for PICs ............................................................... 58 7.1 PICTA and PACER Create a Need for Tax Reform ............................................ 58 7.2 Tailor the Tariff and Tax System to Need and Local Institutional Capacity ....... 59 3. LABOR MARKETS AND PUBLIC SECTOR REFORM: EASING ADJUSTMENT TO TRADE LIBERALIZATION ..................... 65 1. Introduction ..................... 65 2. Factor and Product Markets in PICs: The Backdrop for Trade Liberalization ........ 66 2.1 Rapidly growing numbers of under-employed ..................................................... 67 2.2 Role of emigration ........................................................... 67 2.3 Shortage of Skilled Workers ........................................................... 69 2.4 Public and Private Sectors ........................................................... 70 2.5 Minimum Wages ........................................................... 73 3. Rigidities in other markets ........................................................... 74 3.1 Land ........................................................... 74 3.2 Retail markets ........................................................... 75 4. Public Sector Reform ............................................................ 76 4.1 Resistance to Reform ........................................................... 77 4.2 Private and Social Costs of Reform ............................................................ 78 5. Reducing Resistance to Reform and Easing Adjustment ......................................... 79 5.1 The Economic Argument ............................................................ 79 5.2 The Political Argument - Severance Packages .................................................... 81 5.3 Public W orks Programs ........................................................... 82 5.4 Training or Retraining Programs ............................................................ 83 5.5 Labor Mobility ........................................................... 83 5.6 Credibility of Reforms ............................................................ 84 6. Conclusion ............................................................ 84 References ............................................................ 85 List of Tables Table 1.1: New Agreements Influencing Trade' in the Pacific ...........................................2 Table 1.2: Basic Economic Indicators of Pacific Island Countries .................................. 10 Table 1.3: Main Merchandise Exports, and Tourism and Other Travel Receipts ............ 11 Table 1.4: Intra-PIC Trade (% of total Exports and Imports) .......................................... 16 Table 1.5: PICs - Sources of Imports (% of total imports) .............................................. 27 Table 1.6: PICs - Destination of Exports (% of total exports) ........................................ 27 Table 2.1: Structure of Government Revenue for Fiji, Samoa, Tonga, Vanuatu and Kiribati ............................................................. 35 Table 2.2: Tariff Rates by Broad Categories (%) ............................................................. 37 Table 2.3: Impacts on Imports and Tariff Revenue of a PIC-FTA (based on pre-1998 data) ............................................................. 38 Table 2.4: Schedule for Tariff Reductions under PICTA ................................................ 41 Table A. 1 Fiji Government Revenue ............................................................. 60 Table A.2 Kiribati Government Revenue ............................................................. 61 Table A.3 Samoa Government Revenue ............................................................. 62 Table A.4 Tonga Government Revenue ............................................................. 63 Table A.5 Vanuatu Government Revenue ............................................................. 64 Table 3.1: Basic economic and social indicators ............................................................. 68 Table 3.2: Annual supply-demand imbalance for formal sector employment ................. 68 Table 3.3: Labor conditions in PICs ............................................................. 73 Table 3.4: Utility costs in PICs ............................................................. 76 List of Figures Figure 1.1: Per Capita GDP Growth (1995-2001) .............................................................4 Figure 1.2: PICTA - Distribution of Population and GDP .............................................. 14 Figure 1.3: Estimated Welfare Gains of a FTA with & without Australia and New Zealand (in % GDP) ............................................................. 18 Figure 1.4: Tariff Collection Rates in the PICs ............................................................. 18 Figure 1.5: Welfare Gains to PICs from Full MFN Liberalization (in % GDP) .............. 19 Figure 2.1: Trade Taxes as a Share of Tax Revenue ........................................................ 36 Figure 2.2: Tariff Collection Rate .............................................................. 37 Figure 2.3: Alternative Estimates of the Potential Loss in Revenue of a PIC-FTA ......... 39 (pre-1998 data) .............................................................. 39 Figure 2.4: Fiji - Trends in Revenue .......................... .................................... 43 Figure 2.5: Samoa - Trends in Revenue ................................... .......................... 44 Figure 2.6: Vanuatu - Trends in Revenue .............................................................. 45 Figure 3.1: Share of Public Sector Employment in Paid Employment ............. ............... 70 Figure 3.2: Government Expenditure (in % GDP) ........................................................... 70 Figure 3.3: Wage Bill as a Share of Government Current Expenditure ........................... 71 Figure 3.4: Government Fiscal Deficit (in % GDP) ......................................................... 71 List of Boxes Box 1.1: Key characteristics of PICs ............................................................. 9 Box 1.2: The Melanesian Spearhead Group: Experience with Preferential Trade in the Pacific .............................................................. 12 Box 1.3: CARICOM-Experience of Caribbean Integration . ......................................... 15 Box 1.4: The Cotonou Agreement: ACP and EU Partnerships After Lome . . 24 Box 1.5: SPARTECA - PICs Preferential Access to Australian and New Zealand Markets, and its Erosion ............................................................. 28 Box 3.1: Increasing FDI inflow into the Pacific: Some Generic Advice . . 80 ACKNOWLEDGEMENTS This report was prepared by a team led by Vivek Suri and consisting of Satish Chand, Ron Duncan, John Fallon, and Craig Sugden. Valuable inputs were received from James Crittle and Andrew Proctor. Preparation of the report was supervised by Sanjay Dhar. Overall guidance for the report was provided by Homi Kharas and Klaus Rohland. Peer reviewers for the report were Kyle Peters and Garry Pursell. The topics covered under this report were discussed with finance ministers and senior officials of the Pacific Island Countries at a workshop held in Sydney in February, 2002. This workshop was organized jointly with the Pacific Islands Forum. Comments received from participants at the workshop are acknowledged. Helpful discussions with development partners (Asian Development Bank, Australia, European Union, France, International Monetary Fund, Japan, New Zealand, UNDP, United Kingdom) at the workshop and during the annual Pacific Donors Meeting are also acknowledged. The valuable assistance received from government officials during country visits is greatly appreciated. The analysis in this report seeks to effectively complement earlier work done at the Asian Development Bank, International Monetary Fund, the Pacific Islands Forum, PFTAC, and the World Bank. The team would like to acknowledge the generosity of these organizations in sharing their information. The team has benefited from discussions with officials from these organizations especially Emma Ferguson, David Forsyth, and Jai Kumar (Pacific Islands Forum Secretariat), and Christophe Grandcolas and Klaus-Walter Reichel (PFTAC). Helpful suggestions and comments from Denise Aldous, Peter Bennett, Natasha Beschorner, Laurie Dunn, Gaurav Datt, Sean Duggan, Bernard Hoekman, Maurice Schiff, and Cyrus Talati (World Bank), and Tony Hughes at various stages of the report are also acknowledged. The processing of the report was greatly facilitated with the assistance of Jane Bleakley, Adelma Bowrin, Joeann Buhagiar, Bettina Floyd, Muriel Greaves, Alison Lykissas, and Donna Quartley- Parker. Generous funding received through the AusAID-World Bank Pacific Facility helped partially finance the preparation of this report. EXECUTIVE SUMMARY Preferential Trade Liberalization in the Pacific 1. As a first step towards integrating more closely with the global economy the Pacific Island Countries (PICs) have agreed to enter into a regional trade arrangement (RTA). As a consequence of multilateral and unilateral efforts towards freer trade worldwide the preferential access granted to PICs in developed country markets is being eroded. In recognition of the loss of any benefits from preferential access, the PICs realize the need to become more competitive in global markets. The Pacific Island Countries Trade Agreement (PICTA) is viewed by the PICs as a 'stepping stone' towards this goal. An important reason for the formation of a preferential trade area, as opposed to unilateral or multilateral liberalization, is the perception that trade liberalization should be a gradual process to limit the disruption and adjustment to their economies. As such PICTA is seen as a 'training ground' for PIC businesses and governments, a preparatory step towards more extensive liberalization in the future. PICTA is also a political vehicle - acting as a group with one voice is expected to arrest the political and economic marginalization of PICs in international fora. 2. This report evaluates the strategy that the PICs have chosen towards liberalizing their economies. As a move that represents a shared view among PIC policy makers on the benefits of a more liberalized trade regime, PICTA is a positive step. Trade liberalization is increasingly viewed by the PICs as a strategy to enhance growth prospects in the face of declining per capita incomes in many parts of the region. However, for the 'stepping stone' to not become a 'stumbling block' the strategy has to be carefully managed: wider trade liberalization will have to be actively pursued and policies that create interests which could stall the move towards greater liberalization will need to be avoided. Although greater liberalization may entail greater adjustment, it is important to note that adjustment costs are typically short-term, while the benefits of trade reforn can be expected to grow with the economy. 3. The report also analyses the interaction of trade liberalization with complementary reforms in taxation and the public sector. These reforms while examined here in the context of trade liberalization are important in their own right - their successful implementation would greatly enhance economic efficiency and growth prospects in the PICs. Impact of PICTA 4. PICTA is seen as a natural way to overcome the smallness of domestic markets. It is hoped that the expanded market of nearly seven million would enable fuller exploitation of economies of scale and attract foreign investment. For these benefits to be exploited, integration would have to be deeper than an elimination of internal trade barriers. Trade facilitation measures covering areas such as quarantine and customs i procedures, and product standards and conformance, to be implemented along with PICTA would be crucial in this respect. Agreements on investment regulations, tax policies, service sector liberalization, and labor mobility, which are envisaged for the future, would further deepen the integration. An issue that would remain for PICTA is that even the combined market will remain small for the scale and competition effects to fully operate. 5. PICs are small undiversified economies with similar exports and very little trade between them. Therefore, the estimated benefits of PICTA are very small. But trade diversion which favors the more industrially advanced members is likely to develop. The preferential removal of tariffs on higher-cost imports from member countries, would allow them to displace lower-cost goods that were previously imported from the rest of the world. The government of the importing country would no longer receive the tariff revenue. In this sense, tariff revenue is transferred in large part to the exporting firms in member countries and to some extent to domestic consumers. Besides creating tension between member countries, the benefitingfirms will tend to lobby against the lowering of external trade barriers and the extension of trade liberalization to non-members. In trading blocs comprising small developing countries the same forces that cause trade diversion also lead to income divergence in favor of the more advanced countries. In some cases of trading blocs among small developing countries, such trade and tariff diversion, and income divergence has resulted in the breakdown of the agreements. Trade diversion is more than a theoretical possibility in the Pacific as the experience with the Melanesian Spearhead Group agreement suggests. PICs can avoid these problems by lowering their tariff barriers not only amongst themselves, but also with all trading partners. A credible inquiry mechanism involving all interested parties, perhaps on a region-wide basis, could therefore be established to debate and guide government policies on trade matters. Beyond PICTA 6. PICTA has been formulated under a framework agreement (PACER) which stipulates that negotiations over a FTA between the PICs and Australia and New Zealand are to commence eight years after PICTA comes into force, if not begun before that by mutual consent or triggered by the PICs commencing negotiations over FTAs with other developed countries. When, for example, the PICs enter into negotiations over reciprocal free trade agreements with the EU in 2002-as envisaged under the Cotonou Agreement-this will presumably trigger negotiations with Australia and New Zealand. Three Micronesian countries also have a provision in their Compact of Free Association with the US whereby they must not give more favorable tariff treatment to countries other than the US. Therefore, the widening of preferential trading arrangements beyond PICTA is inevitable. Only the timing, extent and benefits of this widening are uncertain. 7. The formation of FTAs between the PICs and high-income countries-the EU or Australia and New Zealand-will generate much greater benefits for the PICs than PICTA, as modelling work on a FTA with Australia and New Zealand has shown. The increased benefits result from the fact that the high-income countries are usually sources of lower-cost imports than PIC members. However, counteracting these benefits will be ii trade and tariff revenue diversion to the high-income countries. There is also the "hub- and-spoke" problem with respect to investment in overlapping FTAs, especially in the case of the EU which is likely to have FTAs with many developing countries. An incentive is created for investment to concentrate in the country that is common to the various trading arrangements-the "hub" country, in this case the EU-where it gains preferential access to all the "spoke" countries. PICs and the EU 8. In the past, the countries that have benefited most from preferential access to the EU market are Fiji, Solomon Islands and PNG. These countries would have an interest in forming a FTA with the EU in order to retain preferential access. PICs that are included in the UN definition of least developed countries (Kiribati, Samoa, Solomon Islands, and Vanuatu) would benefit from the "Everything But Arms" (EBA) proposal of the EU that would allow duty-free and quota-free access to its market. For these countries there is thus little to be gained from an FTA in goods with the EU. The US compact countries have virtually no trading relationship with the EU. Thus it may be difficult for the EU to have a single FTA with all PICs. In forming regional economic partnership arrangements (REPA) with the EU it would be useful to explore the benefits from covering areas where all PICs have an interest e.g. services like tourism, and the fisheries sector. PICs, Australia and New Zealand 9. Australia and New Zealand figure prominently in the trade and investment patterns of a number of the PICs. Therefore, closer integration with Australia and New Zealand is likely to have a greater pay-off in terms of trade expansion than integration with other regions-except perhaps for the East Asian region which offers complementarity in trading patterns with the PICs. 10. Australia and New Zealand offer considerable scope for deep integration with PICs along the lines of the member countries of the EU. A key potential gain from deeper integration with Australia and New Zealand would be from the free movement of labor and capital. While investment can move with freedom from the PICs to Australia and New Zealand, it cannot move freely in the opposite direction. It is ironic that countries so short of capital hinder its entry through direct and indirect barriers. 11. Labor can move relatively freely from several of the PICs to metropolitan countries such as Australia, Canada, New Zealand, and the United States. However, for the Melanesian group of countries such labor movement and emigration is not available. Further, labor movement in the opposite direction-such as from Australia and New Zealand-is generally tightly controlled throughout the PICs. There appears to be a widespread view in the PICs that employment of labor from high-income countries is a negative-sum game, with such labor displacing local labor. However studies in the region have shown that expatriate labor and local labor are complementary- employment of skilled expatriate labor leads to increased employment of local labor. iii Enhanced labor mobility would be crucial if the regional integration is expanded to include services. 12. Based on experience elsewhere, the formation of RTAs such as the REPAs or with Australia and New Zealand could be beneficial to the PICs, with benefits flowing in the form of technology transfer, cheaper inputs, and the locking-in of economic policy reforms. PACER also provides for the establishment of a program of trade facilitation measures for the island members. Australia and New Zealand are to provide the financial and technical assistance for the implementation of this program. 13. An important constraint to the development of export industries in PICs will be that domestic firms have had little experience in exporting. A proven way of overcoming this hurdle is to form joint ventures with foreign firms that already have established marketing and distribution networks. Providing an encouraging environment for foreign investment to establish joint ventures should therefore be an important objective if trade liberalization is widened to include Australia, New Zealand, the EU, and the United States, as exporting into these major markets will require such assistance even more. The WTO Alternative 14. At present only Fiji and Solomon Islands are members of the WTO. By liberalizing within the WTO framework, the PICs would gain improved market access to the rest of the world without transferring tariff revenue to the EU or Australia and New Zealand or to the United States. The WTO offers the lock-in benefits of an international treaty-probably to a greater extent than a RTA-because it provides both rewards in terms of financial and technical assistance as well as sanctions. Moreover, WTO membership captures all the benefits of technology transfer from integration with other countries, without the problems of hub-and-spoke arrangements. 15. While the transaction costs of negotiating accession to the WTO are high the negotiations to be entered into with the EU, Australia and New Zealand, and the United States would also be extensive and complicated. With prospective membership of multiple trading blocs, complexities arising from the need to reconcile the requirements of the different blocs, the incompatibility of rules of origin, and incompatibility of legal and other institutions is almost unavoidable. Resolving these issues will place a huge burden on the limited skilled resources of the PICs. 16. However, the recent experience of PICs with the WTO accession process has not been positive. Tonga applied to join the WTO in June 1995, Vanuatu in July 1995, and Samoa in April 1998. None of the three has yet acceded to the WTO. Another pervasive problem for small economies within the WTO context is the high cost of maintaining membership, with the need for representation in Geneva to attend up to 50 meetings a year. Hence, consideration has to be given to making WTO membership accessible to small countries at less cost than at present. 17. "Behind the border" legal and institutional reforms in the PICs will be a key component in expanding trade, be it among themselves or with the rest of the world. iv These include reforms in areas such as sanitary and phytosanitary standards, technical standards, investment legislation, competition policy, and property rights. Technical assistance in designing and implementing these reforms in the special circumstances faced by PICs could be the main role played by PACER and Cotonou. Such assistance would also help in the implementation of WTO commitments. Interaction of Trade Liberalization and Tax Reform 18. Tariffs or other taxes on imports are an important revenue source for the PICs, ranging from about 20 percent of tax revenues in Samoa to over 60 percent in Kiribati and Tonga. The reduction in tariffs to take place under PICTA has the potential to erode this revenue. As trade between the PICs is small, the direct loss in revenue will be limited. However, the revenue loss can be expected to rise over time for at least some countries as trade diversion takes place. Moreover as PICTA is widened to include developed country partners, tariff revenue loss would increase for all PICs. As noted, to minimize the adverse impacts of trade diversion, PICs should lower tariff barriers not only amongst themselves but also with all trading partners. This strategy would lead to greater benefits for consumers, but may entail larger revenue loss. But alternative ways of raising revenues are available that could both offset the revenue loss and make beneficial trade more likely under PICTA. Role of the VAT and other tax reforms 19. A key reform often recommended to offset revenue losses from trade liberalization is the introduction of a broad based consumption tax like a VAT. Some PICs already have VATs which were introduced with a substantial reduction in tariffs. An examination of revenue trends in three such PICs, Fiji, Samoa, and Vanuatu shows that the VAT has helped in making up for the revenues lost through tariff reductions. These countries could cope relatively easily with any revenue impacts of PICTA. They could rely on a combination of a small increase in the VAT rate, increasing excise tax rates or introducing new ones, and improving compliance. For a country like Fiji no changes may in fact be necessary. However, all countries would have to consider broader reforms with trade liberalization beyond PICTA. These reforms could include, in addition to the modest ones noted above, taxation of public enterprises, removing exemptions on export income, bringing large scale primary production into the tax net, and a significant effort at improving tax administration to enhance tax compliance and reduce collection costs. These reforms would not only help offset revenue loss but would contribute to enhancing economic efficiency. Improvements likely to be required in VAT administration include faster processing of rebates, especially for zero-rated sectors, improving compliance at ports where most of the VAT is collected, better targeting of larger tax payers and upgrading capacity for conducting audits. 20. Two countries that do not have a VAT and are more exposed to revenue losses are Kiribati and Tonga. These countries could lose in the range of 15-20 per cent of tariff revenues under PICTA and more under wider liberalization. But given their very narrow production base, with very small manufacturing sectors, and low administrative capacity to effectively implement a relatively complex tax, the case for a VAT needs to be carefully v evaluated. In fact, given that a large share of consumption is imported, a tariff is akin to a consumption tax in these countries. The key trade-offfor these countries is the gain in' economic efficiency versus the administrative costs of a VAT. Even in the more diversified PICs with higher levels of administrative capacity, implementation of the VAT has not been without difficulty, and it is not clear how large the efficiency gains have been. Thus for the countries without a VAT, the options of simpler broad based taxes-e.g. a single-stage sales tax, with a mechanism for dealing with taxes paid on inputs by exporters or larger businesses-need to be evaluated. Labor Markets and Public Sector Reform: Easing Adjustment 21. In the late 1990s a number of PICs embarked upon reform programs triggered by actual or impending fiscal crises; some others are gearing up for similar programs. The deterioration in fiscal positions were in turn the result of bloated public sectors, and weak economic management exacerbated in some cases by unfavorable external events. The reforms generally entailed a rightsizing of the civil service, privatization or corporatization of SOEs, and the creation of more conducive business environments. Interaction of public sector reforms and trade liberalization 22. What implications do these public sector reforms have for the current efforts that PICs are undertaking to integrate their economies more closely with the global economy? It has been argued that a large public sector creates distortions in the PIC economies: for example, by acting as a wage leader in the labor market, it leads to wages paid throughout the economy that are well in excess of those justified by productivity levels. These distortions, if left unaddressed, could complicate the adjustment to trade refonn, for which a well functioning labor market is crucial. Similarly, the prices of infrastructure services provided by SOEs or poorly regulated private entities are high in relation to competitors. As a result PIC producers will find it difficult to compete successfully in a liberalized environment. Thus reforms that reduce such distortions could facilitate the adjustment to trade reform. 23. Public sector reforms have not met with expected success. According to a recent review the reforms faltered on three counts. First, the priority given to fiscal stabilization led to a focus on reducing public expenditure through downsizing at the expense of public sector effectiveness. Second, sufficient time and attention was not devoted to addressing the resistance to change that is generally exhibited. Third, insufficient attention was paid to the capacity of the private sector in PICs to respond to public sector downsizing and the liberalization of policy frameworks. 24. Where good opportunities for emigration exist, a reduction in the public sector work force may not lead to a decline in the wage rate. This is because the wage rate in these countries is influenced by the wage which emigrant workers are able to earn in the host country. A downsizing of the public sector could lead to heavy emigration as the experience of Cook Islands, and to a lesser degree Marshall Islands, demonstrates. To the extent that the private sector had developed to service the public sector, the private vi sector could be expected to contract, especially if those who emigrate are the more highly skilled people. 25. The public sector is sometimes viewed as a necessary mechanism for redistributing aid and resource rents in the Pacific. As long as aid flows and resource rents can be expected to remain large, it is argued that there is no need to downsize the public sector. However, the maintenance of a large public or non-tradeable sector is at the expense of the tradeable sectors. Another negative aspect of this form of development is that since the non-tradeable sector is located mostly in urban areas there is overcrowding in towns with attendant environmental and social problems. 26. Once protection is lowered, attempts to maintain previous wage levels will render import competing industries uncompetitive. The same is true of those export industries that have received preferential access or other subsidies. If the environment for the private sector to develop is unfavorable, and emigration possibilities are not available, the displaced labor will either retreat to subsistence or remain un-or under-employed. Improving the business environment therefore assumes great importance. 27. Given this backdrop, and the risk that the private sector response to trade liberalization may be slow to develop, the pace, sequencing and composition of public sector reforns in the specific circumstances faced by individual PICs need special focus. Overcoming resistance to reform 28. Opposition to public sector reform has been witnessed in some PICs, especially those downsizing. Such opposition could also generate or exacerbate resistance to reform in general, including trade liberalization.' Trade liberalization is unlikely to harm (and will likely benefit) the poor and unskilled, the bulk of whom rely on subsistence agriculture for their livelihood. It will threaten high-paying jobs in currently protected sectors and diminish the value of rents accruing to holders of import licenses. This loss ofjobs would add to the labor displaced due to public sector reform, and the large cohort of working age people being added to the ranks of the unemployed population every year. 29. The predicament for policy makers is that while trade liberalization has the potential to create jobs, there may be a time lag between job loss and job creation. The length of lag would depend upon the conduciveness of the domestic environment for investment, and ease of labor mobility. Both of these can be addressed, but resistance to reforn may build in the interim. Thus limiting resistance through compensation mechanisms or other temporary income support programs, or perhaps by adjusting the pace of reforms, may be warranted. 30. Severance packages could be offered to public sector employees to both diffuse resistance and ease their adjustment to reform. The workers typically offered severance packages are likely to be part of the urban middle class. Therefore, in addition to l Adjustment resulting from PICTA and hence the opposition to PICTA is likely to be small. The discussion here should be viewed in the context of fuller trade liberalization which, as argued in chapter 1, may be a better strategy but would entail greater adjustment. vii severance packages, income support programs like public works would be needed to assist people who might temporarily fall into poverty as a result of reform. The job loss of an urban worker could adversely impact dependents who supplement their subsistence livelihood with a small cash transfer from the worker. Income support programs could be helpful for such households. Other mechanisms to ease adjustment costs and reduce resistance to reform include retraining programs, enhancing labor mobility in the region, and improving the credibility of reforns. Implementing programs to ease adjustment costs would place additional demands on PIC budgets in the short-run. In the past, donor assistance has been forthcoming to support reform programs in the PICs. There is a sound case for such assistance to continue especially for well designed and credible reform programs that are strongly owned by the PIC governments, and are broadly supported by major stakeholders. 31. PICTA is expected to be a stepping stone to more complete liberalization. Wider liberalization would potentially bring larger benefits than preferential liberalization. It is, therefore, important that PICTA is implemented in a manner that avoids the creation of interests which could block the move towards greater integration with the global economy. Key to the exploitation of the benefits of greater liberalization would be a conducive environment for business development, and enhanced labor and capital mobility in the region. These factors would also greatly influence the success of public sector reforms that PICs have undertaken or may undertake. The pace and composition of reforms-in particular more complete trade liberalization and public sector reform- will therefore need to be calibrated to the progress on efforts to enhance the business environment, and increase labor and capital mobility. viii I 1. [NTEGRATING WITH THE GLOBAL ECONOMY: TRADE STRATEGIES FOR THE PACIFIC 1. A New Trade Regime for the Pacific 1.1 The idea of a preferential trade arrangement among the Pacific Island Countries (PICs) was first raised in 1971 at the inaugural meeting of what was then the South Pacific Forum, but concrete steps towards a free trade agreement took place only recently. As a result of negotiations, two agreements were endorsed in Apia in June 2001 and formally approved in Nauru in August 2001. The Pacific Agreement on Closer Economic Relations (PACER) is an umbrella or framework agreement that establishes the guidelines for the future development of trade relations among the 14 PIC' members of the Pacific Islands Forum, and with other trading partners such as Australia and New Zealand, the European Union (EU) and the United States. PACER will come into force as soon as it is ratified by seven of the member countries. The Pacific Island Countries Trade Agreement (PICTA) is the first regional trade agreement (RTA) to be established under PACER. PICTA is a free trade agreement (FTA) between the 14 PICs that will allow PICs to retain different levels of external tariffs, but tariffs within the FTA would fall to zero over a 10-12 year period. PICTA will come into force when it has been ratified by six of the 14 member countries. PICs that wish to join after the period allowed for ratification has expired will have to apply as new members. The so-called US "Compact" countries (Federated States of Micronesia, Palau, and Republic of the Marshall Islands) have been given an additional three years to sign the agreement in consideration of the special circumstances of their relationship with the United States. Table 1.1 provides an overview of the new trading arrangements likely to be introduced in the Pacific. 1.2 PACER provides a framework for the gradual development of trade and other economic cooperation among PICs and between PICs and other countries. It acknowledges that different PICs may need to move at a different pace in developing closer integration with other countries. PACER also provides for the establishment of a program of trade facilitation measures for the island members. Australia and New Zealand will provide financial and technical assistance for implementing this program. 1.1 Rationale for PICTA 1.3 Forming a preferential trading area is seen by Pacific Islands policymakers as a natural way to overcome the smallness of their domestic markets. The creation of an IOf the 14 Pacific Island countries 10 are World Bank members. These are Federated States of Micronesia, Fiji, Kiribati, Republic of Marshall Islands, Palau, Papua New Guinea, Samoa, the Solomon Islands, Tonga, and Vanuatu. Cook Islands, Republic of Nauru, Niue, and Tuvalu, are the other four. The 14 PICs, together with Australia and New Zealand, form the Pacific Islands Forum. This report focuses on nine Bank member countries other than Papua New Guinea. 1 Table 1.1: New Agreements Influencing Trade in the Pacific Agreement Status Time Frame PICTA Currently ratified by 4 Tariff reductions to begin countries (Cook Is., Fiji, when agreement enters in to Free Trade Agreement among Samoa, Tonga); needs 6 force. Maximum tariff would 14 PICs. ratifications to come into force be 20-25 percent once PICTA begins. Tariffs are to be Expected to be ratified in 2002 reduced to zero by 2010-2012 in a phased manner. Additional 3 years for US Compact countries. PACER Currently ratified by 6 FTA Negotiations with countries (Australia, Cook Is., Australia and NZ to begin Framework Agreement under Fiji, New Zealand, Samoa, either: which PICTA is signed. Tonga); needs 7 ratifications Provides for gradual to come into force 8 years after PICTA comes development of trade and into force. other economic cooperation Expected to be ratified in 2002 among PICs, and between Or earlier by mutual consent PICs, Australia and New Zealand. Includes financial Or triggered as soon as PICs and technical assistance for enter into free trade trade related activities. Would negotiations with other involve negotiations over developed countries (e.g. EU FTAs at a later date. see below). Cotonou Accord Ratified by all PICs Negotiation on REPAs to begin in 2002 and must be Successor agreement to Lome concluded by end-2007. Until between the EU and ACP then EU will continue Lome countries - includes all 14 preferences. PICs. EU to form regional economic partnership agreements (REPA) with PICs. This could be one REPA with all PICs or different ones with groups of PICs, or individual PICs. REPAs would involve negotiation of free trade areas with PICs. 2 expanded regional market of 7 million consumers would enable firms to take fuller advantage of economies of scale. The small size of domestic markets also acts as a barrier for attracting investment. It is hoped that the creation of a larger regional market with zero internal tariffs will encourage investment, both domestic and foreign. 1.4 PICTA is a response to the changing global aid and trade regimes. With the on- going liberalization of world trade and investment resulting from unilateral, regional and multilateral initiatives, the margin of preference accorded to PICs in developed country markets is being eroded (see Section 5 below). In recognition of the loss of any benefits from preferential access, PICs realize that they have to become more competitive in global markets. 2 PICTA is seen as a first step towards this goal. In the post-Lom6 period, the creation of PICTA was also viewed as being able to greatly facilitate the development of a reciprocal trade agreement or regional economic partnership arrangements between the PICs and the EU. 1.5 PICTA is also a political vehicle that can help arrest the political and economic marginalization of the PICs. The expectation is that the more the region acts as a group with one voice, the more political influence it will have in international fora. 1.6 A major reason for the formation of a preferential trade area, as opposed to unilateral or multilateral liberalization, is the perception that trade liberalization should be a gradual process to limit economic disruption and adjustment. This was also the reason for not including Australia and New Zealand in the preferential trading area at this stage. The PICs recognize that a limited PICTA will not by itself overcome the challenges of small size and the changing global trade and aid regimes. For this reason, PICTA is viewed by the PICs as a "stepping stone" towards their more complete integration into the international economy, a preparatory step towards more extensive liberalization in the future. It is hoped that PICTA will provide the PICs with experience in the negotiation and operation of a FTA, and will help to create a common basis for the PICs' negotiations with other trading partners. PICTA is seen as a "training ground" for the PIC economies, ready for further integration in the future. Businesses will become used to the idea of outside competition, and some governments will have begun implementing reforms needed in preparation for more extensive trade liberalization, for example tax reform. 1.7 PICTA is a positive step that represents a shared view among PIC policymakers on the benefits of trade liberalization. In addition to reflecting a response to changing global trade and aid regimes, trade liberalization is increasingly viewed by the PICs as a strategy to enhance growth prospects in the face of declining per capita GDP in many parts of the region (Figure 1.1). The opportunities created through a more open trade and investment environment would be an important component of a program to tackle rising youth unemployment in the PICs. The challenge for policymakers would be to build on the dynamic created by PICTA and ensure that the 'stepping stone' does not become a 2 For example, the European Commission (EC DG-VIII 1999) calculated that by the year 2000 Lomd Convention preferences would have been only 2.9 percent on the basis of the MFN trade regime and 2 percent on the basis of the GSP regime, nearly one-tenth of their current level in some cases. 3 'stumbling block.' This chapter evaluates the strategy that PICs have chosen to more closely integrate themselves with the world economy. It argues that reducing trade barriers with all trading partners simultaneously with PICTA would bring greater benefits than PICTA alone. It also recommends that the liberalization process be carefully managed to avoid policies that foster vested interests which could stall the transition to more extensive liberalization. Although greater liberalization may entail greater adjustment, it is important to note that adjustment costs are typically short-term, while the benefits of trade reform can be expected to grow with the economy. Figure 1.1: Per Capita GDP Growth (1995-2001) 4% - 2% 0%- -2% -r ' t ' / o -4% -6% -8% 1.8 The rest of this chapter is organized as follows. Section 2 analyzes the potential impacts of PICTA and evaluates the challenges that PICs face in exploiting the expected benefits from PICTA. These are divided into non-economic and economic considerations. The standard non-economic reasons for the formation of a RTA may not have much force under PICTA, but could potentially be more important under RTAs with Australia, New Zealand and the EU. Economic impacts of PICTA can be usefully categorized as 'scale and competition' effects and 'trade and location' effects. The discussion in Section 2.2 shows that to exploit the scale economies offered by a combined market, integration would have to be deeper than simply an elimination of tariff barriers. The program of trade facilitation to be implemented along with PICTA would be very important in this regard. Trade and location effects are also analyzed in Section 2.2. The crucial trade impact is trade diversion and the associated transfer of tariff revenues to partner countries. The same forces that are behind trade diversion could lead to relocation of industries to the more industrially advanced PICs and cause incomes to diverge among member countries. These impacts are perhaps the greatest challenge for PICTA. Both can, again, be countered if PICs reduce their tariffs not only amongst themselves, but also externally with the rest of the world. Section 3 reviews earlier studies that have 4 attempted to quantify some of the economic impacts of PICTA. As expected, PICTA alone has only a small net benefit. However, net benefits rise substantially when Australia and New Zealand are included in the FTA. Some key features of PICTA as negotiated and their implications are discussed in Section 4. The importance of a credible public inquiry process that involves all interested parties to debate government policy on trade issues is stressed. Section 5 looks at the issues beyond PICTA that PICs will have to grapple with on their way to integrating more closely with the global economy. The PICs relationship with their key developed country partners is discussed. One consequence of PACER and the PICs relationship with the EU under the Cotonou agreement is that the widening of PICTA to include developed countries is almost inevitable. Section 5.1 discusses the future preferential trade relationship with the EU and section 5.2 with Australia and New Zealand. There are benefits to be had for PICs especially if these agreements are extended beyond goods to include services and labor mobility. But the process of negotiating preferential agreements could be demanding on the limited administrative resources available to the PICs. An alternate route to international integration is multilateral MEFN liberalization under the WTO. Section 6 examines the arguments for and against the WTO route in the PICs context. While there are many benefits, the WTO accession process is costly for small economies and PICs recent experience has not been favourable. It is therefore suggested that ways to smooth the WTO accession process for PICs needs to be explored. Section 7 concludes. 2. Analyzing the Impacts of PICTA 2.1 Non-Economic Impacts of PICTA 1.9 It is widely recognized that most RTAs are undertaken not for their direct economic impact but for political or indirect economic reasons. The largely non- economic objectives suggested are intra-regional or extra-regional security, enhanced bargaining power in international fora, greater cooperation on other shared issues, and as a commitment and 'locking-in' mechanism for policy reform. How relevant are these reasons to the formation of a RTA between the PICs? Security 1.10 A take-over by external forces does not appear to be a high probability threat, nor is there a significant risk of intra-regional conflict. At any rate, a security arrangement with a large, high-income country would appear to be a better way to protect against external physical threats than a trade agreement among the PICs. The PICs do feel some degree of physical threat from global warming. However, they already cooperate intensively and effectively in lobbying against this perceived threat through the Pacific Islands Forum. 1.11 Through frequent meetings at different levels of government, the Pacific Islands Forum also serves as a mechanism for developing cooperation and trust between the PICs on other issues. Thus, it acts as a means for counteracting any tensions that may be building between the countries. 5 Policy Lock-in 1.12 Some RTAs have proven to be effective means of locking-in economic reforms. The international obligation under a RTA protects governments against pressures from domestic interests to overturn the reforms. This provides for some consistency and predictability in policy from one government to the next and, therefore, gives investors greater confidence in the stability of policies. As Duncan (1994) argued, the EU and the North American Free Trade Agreement (NAFTA) have been effective in locking-in good policies adopted by new members: Spain and Portugal in the case of the'EU, and Mexico in NAFIA. Indeed, these RTAs are much more than trade agreements because of the underlying economic compact between the members. In return for the relatively less- developed countries' promise to maintain good policies, the other members offer support in times of crisis. 1.13 It has been suggested that the PICs, Australia and New Zealand could constitute such a wider economic pact as a means of supporting better policy and institutional performance in the PICs. Also, the Cotonou Agreement between the EU and the ACP (Africa, Caribbean and Pacific) countries stresses the importance of good macroeconomic policies and the role of peer pressure in helping ACP countries maintain sound policies, and is much more than a trade agreement. Incentives through performance based access to development funds are explicitly included. Although PICTA is an international obligation that governments can point to in resisting pressures to overturn reforms, it is not able to provide the same kind of incentives for sustaining good policies as the Cotonou agreement, or agreements with Australia and New Zealand: to lock in good policies, sanctions and rewards are needed. To the extent that the EU, Australia and New Zealand were able to enforce sanctions in the event of poor performance and, equally, reward strong performance through increased access to development assistance, RTAs with them would offer greater potential to lock-in good policies. 1.14 It is relevant to note that the WTO also provides a means of locking-in trade reform through governments committing to an international obligation with accompanying sanctions and rewards. The binding of tariffs is an explicit commitment to better policies that is enforced through the WTO by sanctions. In return, there is the offer of various forms of technical and financial assistance to improve capacity in human resources, and in developing better policies and institutions. However, as argued below the recent experience of some PICs with the WTO accession process has not been favorable. 2.2 Economic Impacts of PICTA 1.15 The economic effects of PICTA can be divided into two types: 'scale and competition effects' and 'trade and location effects'.3 These are discussed in turn below. 3 The conceptual discussion in this section draws upon the following: World Bank, 2000, Trade Blocs. A Policy Research Report. Washington, D.C. 6 Scale and Competition Effects 1.16 A common argument in favour of PICTA has been the potential exploitation of economies of scale that an expanded regional market would offer. The simple image this may convey is a large factory serving the entire region. However, this would not be beneficial for the PICs as such a factory would likely be a protected monopoly with high prices for its products. If one of the benefits from the larger regional market is to be lower prices then the key is to secure effective competition. Individually, each PIC market is too small and more competition within it would mean smaller factories which are unable to exploit scale economies. The regional market enables both increased competition and a larger average scale. Lower prices for consumers will result from the combined effect of increased competition and scale. Removing tariffs on imports from partner countries is one component for securing the gains from competition and scale effects. Also required would be supporting policies that enable all firms to have real penetration in all countries in the region. In other words, regional integration would have to be deep, as for instance in the EU. Trade facilitation measures covering areas such as quarantine and customs procedures, and product standards and conformance, to be implemented along with PICTA would be crucial in this respect. Agreements on investment regulations, liberalizing the service sector, and labor mobility, which are envisaged for the future, would further deepen the integration.. An issue that would remain for PICTA is that even the combined market may remain too small for the scale and competition effects to fully operate. Foreign Direct Investment (FDI) 1.17 A related matter is whether, and under what conditions, a larger market attracts FDI. A foreign investor will weigh the costs of supplying a product from outside the region against the cost of producing at a plant located within the region. If the product is supplied from outside then tariff and other trade barriers are an important component of cost. If these barriers are high then the incentive to undertake FDI could be created mainly by the desire not to pay tariffs to the local government. Such "tariff jumping" FDI can reduce real income as the costs of producing locally may be greater than the cost of imports. On the positive side, there is evidence that FDI can play a role in stimulating production in related local industries, in transferring technology and in raising productivity of neighboring firms. However, as with the scale and competition effects even the combined PICTA market may be too small to attract large amounts of FDI. Also in order for the potential investors to actually view the region as one seamless market, integration would have to be deep. 1.18 While attracting FDI, PICs would need to guard against tax- competition. A foreign investor can gain tariff-free access to the entire market by investing in any country that is part of PICTA. This may create a tendency for governments to compete against each other in offering tax and other concessions in order to attract the foreign investor. Doing so might set off a tax "race to the bottom." To prevent this from happening, tax policies for foreign investment may need to be harmonized or PICTA extended to control such tax concessions and other means of subsidizing investments. 7 Trade and Location Effects 1.19 If they can be exploited, the competition and scale effects will accrue to the region as a whole. Trade and location effects, on the other hand, relate to transfers from one part of the region to another. Trade creation and diversion 1.20 The trade impacts of RTAs are normally measured by weighing the benefits of trade creation against the adverse effects of trade diversion. Trade creation occurs when lower-cost partner country imports displace higher-priced domestic production, and thereby decrease costs of production and consumer prices within the country. Trade diversion happens when lower-cost imports from outside the bloc are displaced by higher-cost imports from within the bloc. 1.21 Generally, trade creation and welfare benefits are greater the higher the trade barriers being reduced and the larger the reduction in barriers; the larger the partner countries; the larger the number of partners; the more diversified the economies of the partner countries; and the closer the partners' prices are to world prices or the more competitive they are compared to countries outside the bloc (Laird 1999). Benefits are also greater when the partner country is not only large and rich, and has a complementary trading pattern, but is also fast-growing. 1.22 It should not be surprising, therefore, that the estimated gains from a RTA among the PICs will be tiny. They are all small economies with low levels of diversification and similar trade pattems (Tables 1.2 and 1.3). The combined population of PICTA countries is only 7 million, and intra-trade accounts for only 4 percent of total trade of all countries. A brief overview of PIC economies is provided in Box 1.1. 1.23 The following example clarifies the trade diversion concept. Suppose the home country can import a can of beer from a potential partner country at $1.20 per can and from a non-partner country at $ 1 per can. If the tariff on both is 40 cents per can, the price paid by the home consumer would be $ 1.60 and $ 1.40 for a can from the partner and non-partner respectively. Consumers would thus prefer to buy from the non-partner. Now suppose that due to a preferential trade agreement the duty is reduced to zero for the partner, but not for the non-partner. The price for the partner now declines to $1.20 and the consumers shift from the non-partner to the partner country. While for the consumer the price is 20 cents cheaper, the government loses 40 cents in revenue per can. The net effect on the country is 20-40 = -20 cents, i.e., a loss of 20 cents. The price charged by the partner need not fall to $1.20 - as long as the price is less than $1.40 the partner would out compete the non-partner. If, say, the price fell to $1.30, the consumer would gain only 10 and the loss to the country would be 30 cents. 8 Box 1.1: Key characteristics of PICs The small size of the domestic market interacts with a narrow resource base to make PICs relatively undiversified and highly open on the trade account, with openness being expressed through a high share of imports to GDP and, in several cases, a high share of exports to GDP. The remoteness and isolation of these countries from major centers of trade and commerce makes it more difficult to compensate for the limited domestic markets by turning to world markets. Transport costs are high because of small cargo volumes. The ratio of freight and insurance debits to import costs (an indicator of transport costs) is 20 percent or higher for PICs compared with the median value of 14 percent for all developing countries. The constraints of a narrow resource base and small size are most apparent in 'advanced' manufacturing activities such as capital goods, metal working, and intermediates. These activities are characterized by economies of scale in production, product development and R&D. The share of manufacturing in GDP in PICs is generally under 10 percent. Manufacturing is thus limited to basic small scale 'low technology' activities like food processing and breweries which are not greatly impacted by market size. In Fiji the share of the manufacturing sector is 15 percent because of the sugar and garment industries, and of a similar magnitude in Samoa with an automotive wire harness manufacturer playing a key role. In the absence of a large manufacturing sector PICs are dependent on the primary and tertiary sectors of the economy as a source of output, employment, and foreign exchange. In the agricultural sector a large share of production is in the form of smallholder operated, semisubsistence, household enterprises growing chiefly root crops and garden vegetables. In Fiji agriculture, especially sugarcane, is organized more along commercial lines, although the subsistence sector remains important. Large scale agriculture comprises oil palm, coconut, cocoa, and coffee plantations, and beef cattle (in Vanuatu) with the bigger and more capital intensive operations wholly or partially foreign owned. The sharp fall in prices of traditional PMC export crops has stimulated producers to move into nontraditional crops to exploit niche markets in squash pumpkin, vanilla, melons, and coconut cream. The contribution of the tertiary sector ranges from 55 to 60 percent in Fiji, Samoa, Tonga, and Vanuatu, while in FSM, Marshall Islands, and Kiribati it is greater than 75 percent. In the tertiary sector the two important activities are public services and tourism related services. Tourism's key role as a foreign exchange earner in the PICs is illustrated in Table 1.2. The economic contribution of the fisheries sector comes from subsistence activities employing traditional methods, commercial fishing, and downstream industrial processing. National accounts statistics show a contribution ranging from about 2 percent in Fiji to about 9 percent in Marshall Islands, but these are greatly understated as they fail to take full account of artisanal and subsistence production. For example, in Kiribati where commercial fishing contributes about 3.5 percent to GDP, the share of subsistence fisheries is estimated at nearly 9 percent. Fish exports are important for a number of PICs (Table 1.3). Most PICs also receive access fees from licensing of Distant Water Fishing Nations (DWFNs) vessels. Total revenues from access fees varies significantly across countries reflecting the great variation in the distribution of offshore fish resources. For the Micronesian countries, fishing fees have in the recent past contributed amounts equivalent to 50 percent to 100 percent of revenues raised from taxes. 9 Table 1.2: Basic Economic Indicators of Pacific Island Countries Population GDP GNI per capita Exports to GDP Imports to GDP ('000) (US$, mil.) (US$), ratio2 ratio2 Fiji 810 1647 2160 28 38 FSM 118 229 2110 7 61 Kiribati 91 43 950 15 79 Marshall Is 52 94 2210 7 73 Palau 19 117 7120 10 79 Samoa 170 237 1420 7 46 Solomon Is 430 277 600 48 51 Tonga 100 159 1630 6 42 Vanuatu 197 226 1100 11 36 PNG 5130 3476 670 57 42 Cook Is3 18 78 4355 7 55 Tuvalu3 10 13 1295 1 51 1. GNI per capita for 2000 using World Bank Atlas method. GDP per capita for Cook Islands and Tuvalu. 2. Average of latest three years for which data are available. 3. Not member countries of the World Bank 1.24 Another way of looking at trade diversion is to ask where do the 40 cents that the government loses go? Twenty cents goes to the higher cost of producing in the partner country relative to the non-partner. The remaining 20 cents is divided between the consumers in the home country and exporters in the partner country. In the above example if the price charged is $1.30 then consumers and partner country exporters both gain 10 cents each. 1.25 It is important to note that trade diversion can occur only if the country has a tariff on imports from the rest of the world, and that the cost of trade diversion cannot exceed the height of this external tariff. In the example, if the external tariff was initially low the loss of tariff revenue would be small, and if the external tariff were cut, the switch in supply-source would not occur. A crucial policy implication is that to avoid the costs of trade diversion member countries of PICTA should lower external tariffs as much as possible. 1.26 There may also be a hidden cost associated with trade diversion. One by-product of trade is knowledge: firms learn from their trading partners. There is evidence that a 10 Table 1.3: Main Merchandise Exports, and Tourism and Other Travel Receipts* Main Merchandise Exports Tourism and Other Travel Receipts (% total merchandise Exports) (% exports of goods, services & income) FSM Fish 79.7 24.3 Banana 1.1 Copra 0.7 Fiji Garments 27.8 22.5 Sugar 21.7 Gold 6.5 Fish 5.6 Kiribati Copra 49.8 3.0 Fish 5.3 Marshall Is Fish ** 8.2 Crude coconut oil 18.5 Copra cake 15.7 Pet fish 5.9 Samoa Fish 53.4 50.2 Garments 9.5 Copra 8.1 Coconut Cream 7.6 Beer 5.0 Coconut Oil 4.7 Solomon Is*** Timber 40.5 3.0 Fish 26.8 Palm oil 12.8 Copra 6.4 Cocoa 4.0 Tonga Squash 50.7 28.4 Fish 32.6 Vanilla 5.2 Root Crops 4.3 Copra 0.9 Vanuatu Copra 37.3 33.1 Kava 14.5 Timber 11.0 Beef 10.2 *Average for latest 3 years for which data are available. ** Data for 1997 showed Fish exports of 79%. * Data up to 1999. Exports of palm oil have now ceased. 11 country's knowledge benefits are directly related to the stock of knowledge of the trading partner, with the stock of knowledge measured by the accumulated investment in research and development. Hence, if a PICTA member diverts its trade from a richer non- PICTA country with a large knowledge stock, to another PICTA member with a much smaller knowledge stock, it will reduce its learning. 1.27 In sum, income redistribution takes place within FTAs as the result of trade diversion. When tariffs are no longer levied on imports from the other members, imports are diverted from the rest of the world to the higher-cost member country firms. The government of the importing country no longer receives the tariff revenue on these imports. In this sense, tariff revenue is transferred in part to the exporting firms in member countries and in part to domestic consumers. The income redistribution gives firms benefiting from it an interest in arguing against the lowering of external trade barriers or against the extension of- the trade liberalization. This form of income redistribution takes place in all RTAs and once again it can be reduced or eliminated by lowering tariffs for all trading partners. An example of trade creation and trade diversion in the Pacific context is provided by the experience under the Melanesian Spearhead Group agreement (Box 1.2). Box 1.2: The Melanesian Spearhead Group: Experience with Preferential Trade in the Pacific Vanuatu' s experience under the MSG Trade Agreement and related bilateral trade agreements illustrates the potential impact that PICTA could have on regional trade. The MSG Trade Agreement is a preferential trade agreement put in place between Papua New Guinea, Fiji, the Solomon Islands and Vanuatu as a first step towards a regional free trade agreement. The Melanesian countries account for the majority of trade between the PICs and full trade liberalization amongst the member countries of the MSG Trade Agreement would have seen an outcome very similar to that targeted through PICTA. But tariffs have only been removed gradually since the agreement's inception in 1993 and bilateral trade agreements have taken precedence over the provisions of the MSG Trade Agreement. So trade between the four countries is yet to be completely liberalized. There has been a reluctance to expose local producers to competition and this meant very little trade creation. Instead, member countries tended to limit tariff concessions to products they did not produce. This saw MSG products displace non-MSG products, and the main consequence of the early years of the agreement was found to be trade diversion and a net loss of revenue. Further, when the Solomon Islands was provided tariff-free access to Vanuatu for its tinned fish, the import price rose by almost as much as the tariff fell, such that there was little change in the price to consumers. There was also evidence that Vanuatu exporters of fresh beef increased their prices of sales to the Solomon Islands as tariffs fell. This can be seen from Figure B2. 1 that shows a large increase at the commencement of the MSG Trade Agreement in the unit price of Vanuatu's beef exports to MSG partners relative to the unit price on exports to Vanuatu' s main buyer of beef, Japan. While the difference has subsequently narrowed, the unit price remains relatively higher for MSG trade indicating than Vanuatu exporters may have continued to capture some of the benefits of the tariff preference. Both exports from Vanuatu to MSG partners and Vanuatu's imports from MSG partners have increased over the period of the trade agreement. More detailed data show that producers in Fiji and Papua New Guinea have displaced non-PIC suppliers. Much of this displacement has been in processed food. 12 Figure B2.1: Vanuatu's Beef Exports 400 Unit price of beef exports to MSG partners Exports to MSG partners 300 so -r:S|w/ 40- 200~~~~~~~~~~~~~~~~ Unit price of beef 30- ~2D 100~~~~~~~~~~~~~~~~~~~~~~~1 : 1OO apons to Japm a~~~~~~~~~~~t 0 0 1993 1995 1997 1999 1993 1995 1997 1999 products, mainly tinned meat, fish and noodles, but it is also evident in some building products such as cement and plywood. This displacement is probably an example of trade diversion. But there are also signs of trade creation. Some local producers have lost market share to exports from the MSG partners, notably the local producers of tinned meat (who had previously largely shared the market with non-PIC suppliers), ice cream and plastic products. In the case of ice cream, imports from Fiji have been retailed at around half the price of the higher quality local product and substantially reduced sales of the local product. In response, the local producer has cut back staff, is exploring options of reducing the quality for the local market while exporting more of its higher quality product. This is an illustration of the type of adjustment that can be triggered by trade reform and over time can be expected to improve the efficiency of the economy. Almost all of Vanuatu's additional exports under the MSG trade agreement have been of fresh and frozen beef. Earlier, Japan was the main market for Vanuatu's beef. The MSG partners offer a lower quality and smaller market than Japan and a focus on these markets may ultimately hinder the development of the industry in Vanuatu. For example, the ability to sell products locally may have delayed moves by the industry for organic certification of their products. This would have helped in accessing non-PIC markets and raising the price received by exporters. Vanuatu's beef exports are an example of trade diversion rather than trade creation. Figure B2.2: Vanuatu's Trade with MSG Partners 8,000 - 12 Share of Irnports exports 6,000 - 8 4,000- 4- 2,000- Exports~ ~ ~~mprt 0 0 - . - 1994 1996 1998 2000 1994 1996 1998 2000 *A lowering in tariffs in the MSG partners would have contributed to this narrowing of the price differential. 13 Location Effects: Income Convergence or Divergence under PICTA 1.28 Experience has shown that while a FTA between high-income countries or between high-income and low-income countries tends to lead to income convergence within the bloc-as has happened in the EU and in the NAFTA- a FTA made up entirely of low-income countries tends to lead to income divergence i.e., the income gap between countries grows larger. The general principle at work is that countries in the bloc with comparative advantage closer to the world average (i.e. ones that are relatively more industrially advanced) tend to do better as the result of trade diversion. The collapse of the East African Community (EAC) in 1977 was determined to be the result of Kenya doing much better in developing its industrial sector than Tanzania and Uganda. Prior to the formation of the EAC, Kenya was more industrially advanced than the other members. The same problem has soured the development of the customs union among Caribbean states, with Trinidad and Tobago accounting for some 75 percent of intra- regional exports (see Box 1.3 on CARICOM). 1.29 The agglomeration effect is another reason for relocation of economic activity. As economic centers start to develop, economic activity tends to agglomerate or cluster in these locations. This extends the advantage of locations that had a head start. PICTA too would cause a relocation of economic activity within the region. As can be seen from Table 1.2 and Figure 1.2, Papua New Guinea and Fiji are much larger than the others in terms of population, economic activity and trade. They are also more diversified in terms of manufacturing and service sector activities. It is therefore likely, though not necessary, that a number of activities stimulated by PICTA will be located in these countries. Figure 1.2: PICTA - Distribution of Population and GDP 80% - 70% - 60% - U Population Share 50% - ,. GDP Share 40% - 30% - 20% - 10% ,1 0% PNG Fiji Others 14 Box 1.3: CARICOM-Experience of Caribbean Integration The Caribbean community (CARICOM) was formed in 1973 with the aim of establishing a common market for goods with a common extemal tariff (a customs union), and cooperation in areas such as agriculture, industry, education, health, culture and sports. Cooperation was also envisaged in extemal economic relations. Countries may be members of the Community without being members of the common market. In 1989, CARICOM Heads of govemment decided to move towards a single market, allowing the unrestricted movement of goods, services and factors of production, including labor. It was also decided to work towards the establishment of a single currency within a monetary union. However, progress towards all of these goals has been slow. Full liberalization of trade has not been achieved. Exceptions continue with respect to agricultural products and a list of sensitive items produced in the less developed countries. Some non-tariff barriers and special trade taxes have also been retained. The adoption of the common extemal tariff (CET) is not comprehensive, mainly because of the fiscal implications of removing trade taxes, and the slow implementation of altemative forms of taxation. Over a decade after the agreement to form a single market, the protocols for its implementation have been negotiated and ratified but govemments have not yet given effect to these measures. The impact of the liberalization of trade between the members has been significant, but unbalanced. Intra-CARICOM exports as a share of total exports has doubled over the past decade, from 12 percent to 23 percent. However, Trinidad and Tobago alone accounts for about 75 percent of intra- regional exports. Trinidad and Tobago has captured market share because of its more diversified production base. On the import side, Jamaica absorbs about 40 percent of intra-regional imports, while only accounting for about 4 percent of intra-regional exports. But while the liberalization of intra-regional trade has focused on trade in goods, the services sector is by far the largest part of the economy of these countries-with services averaging around 60 percent of GNP. Tourism is the dominant services industry, but exports of financial, professional and business services, as well as entertainment, are growing. In order to liberalize trade in services, attention has to be given to free movement of labor-which is not the case with goods trade. Cooperation in various regional activities have seen both successes and failures. Intra-regional cooperation in education, health, disaster preparedness, and police and security have been successful. Production integration efforts in the fields of agriculture and manufacturing, involving the establishment of regional State enterprises have not succeeded. Cooperation in extemal economic relations is seen as perhaps the most successful part of CARICOM. Negotiation of bilateral relations with the EU over the Cotonou Agreement, promotion of the interests of small states within the WTO, and negotiations with other regional blocs and individual countries are seen as successful examples of cooperation action in extemal relations. Source: Alister McIntyre, "Challenges Facing Caribbean Integration," paper presented at the Forum on Intal: 35 years of Commitment to Regional Integration, held in Buenos Aires, Argentina, 27-8 November, 2000. * CARICOM has 15 member countries, Bahamas, Barbados, Belize, Guyana, Haiti, Jamaica, and Trinidad and Tobago, plus the eight members of the Organization of Eastern Caribbean States (OECS). 15 3. Quantitative Estimates of the Economic Impact of PICTA 1.30 Quantitative analysis of the benefits and costs of a PIC-FTA has confirmed that the net gains would be relatively small. Intra-PIC trade is presently very limited (Table 1.4) and is not likely to increase very much as a result of the preferential reduction of trade barriers. In a study for the Pacific Islands Forum Secretariat, Scollay et al (1998) showed that the net welfare gains, in terms of real consumption, would be around A$5 million a year. The gains would be larger if the bloc's trade barriers against the rest of the world are also lowered. It was estimated that as a result of the formation of the bloc, trade among the PICs increases as would exports to the rest of the world. However, tariff revenues would decline in all countries, with the decline being larger the higher the country's tariffs and the greater the reduction in tariffs facing the rest of the world. Table 1.4: Intra-PIC Trade (% of total Exports and Imports) Country Exports to PICs Imports from PICs Fiji 11.7 0.4 FSM 0.1 0.0 Kiribati 3.9 10.0 Marshall Is 0.1 1.0 PNG 0.6 0.7 Samoa 1.9 9.7 Solomon Is 2.2 3.5 Tonga 4.3 8;0 Vanuatu 4.5 9.3 Source: Filmer and Lawson (1999, p. 11) and Bank staff estimates. 1.31 The quantitative economic analysis (in the form of computable general equilibrium modeling) has four shortcomings that bias the benefits of the FTA downwards. First, such models are constructed on the basis of existing activities and therefore cannot predict new activities that may come into being because of the lower input costs and lower exchange rates resulting from trade liberalization. Second, such models do not capture the dynamic benefits of trade liberalization. For example, increased productivity cue to greater competition from imports, and the new ideas that may be brought into countries with the larger volume and variety of imports, is not captured. Third, the study did not include the liberalization of services trade between the PICs. Services are generally highly protected and, with liberalization, trade in services could grow significantly as services trade is not as limited by the physical remoteness of the PICs as is merchandize trade. Fourth, abolition of non-tariff barriers was also not included in the modeling exercise. 1.32 On the downside, trade diversion may create activities that will not be competitive when tariff reductions take place later within a wider FTA or through multilateral liberalization as a result of WTO membership. These activities are not necessarily 16 captured in the modeling exercise and the distribution of income after the formation of the FTA could therefore be more biased towards the better-endowed countries than is shown by the model results. 3.1 Quantitative Estimates of Including Australia and New Zealand in the FTA 1.33 Scollay et al (1998) note that the inclusion of the French and US Pacific territories in the PIC-FTA would more than double the size of the market and would therefore increase the benefits of preferential trade liberalization. Though such an expansion of the PIC-FTA was not modeled, it is plausible to expect that the benefits would still be small in absolute terms and would not overcome any of the problems mentioned above. 1.34 An expansion of the PIC-FTA to include the two high-income members of the Pacific Islands Forum, Australia and New Zealand, as allowed for under PACER, has also been modeled using the same technique (CIE 1998). The estimated welfare gains to the PICs (not including the benefits to Australia and New Zealand) total A$205 million annually, compared to the A$5 million gain from the PICTA. The additional gains largely stem from lower-cost imports displacing higher-cost domestic production in the PICs. Welfare changes measured in terms of real consumption under the two FTAs are shown in Figure 1.3. It needs to be stressed that the models used suffer from data deficiencies, and employ a highly aggregated representation of the economy. For this reason, it is important to focus on the relative magnitudes and direction of the results, rather than on absolute values generated by the model. 1.35 PICs could be divided into three categories with respect to the welfare gains from the expanded FTA (including Australia and New Zealand) and the loss in tariff revenue: (i) those where there is a large gain in welfare because tariffs are reduced from a high pre-existing level but where it would be relatively easy to replace lost tariff revenue because they have a VAT (e.g. Fiji and Papua New Guinea); (ii) those with a small welfare gain and little loss in tariff revenue because existing tariffs are low or trade with other members of the FTA is low (e.g. Federated States of Micronesia and Palau); and (iii) those countries with a significant gain in welfare but where there is not a VAT-type tax and major tax reforms will be required to replace the lost tariff revenue tax and major tax reforms will be required to replace the lost tariff revenue (e.g. Marshall Islands, Kiribati, Solomon Islands, Tonga)4. 1.36 The models used to estimate the impacts of the FrA employ collected tariff rates (the ratio of tariff revenues to total imports) to measure trade barriers. These are typically much lower than what would be expected on the basis of published tariff schedules due to widespread duty exemptions and under-collection. Since the modeling studies were conducted, there have been changes in this ratio (Figure 1.4). In the PICs where collected tariff rates have declined due to reductions in actual tariffs, expected welfare gains from an FTA (with or without Australia and New Zealand) would be larger. 4Both Samoa and Vanuatu were included in this category in the study, but they both now have a VAT. 17 Figure 1.3: Estimated Welfare Gains of a FTA with & without Australia and New Zealand (in % GDP) 3.5 U w/o Aus. and NZ 2.5 - | with Aus. and NZ 1.5. 0.5 v- io:~~~~~~~~~0 -1.5 - -2.5 Note: For Samoa the scenario in Scollay et al. (1998) incorporating a 25% lower tariff is presented as it is closer to Samoa's current rate. The scenario with a higher tariff rate estimated a welfare reduction of 0.68% of GDP in the FTA without Australia and New Zealand. For all countries the results of the scenario that assumes the existence of unemployed resources are reported. Source: Scollay et al (1998) and CIE (1998). Figure 1.4: Tariff Collection Rates in the PICs* 35% - 30% El Old (Used in model) 25% * Recent 20%- 15% 1 10% 5% Source: CEE (1998) and Bank staff estimates. *Recent refers to 1998-2000. 1.37 In the models, as tariff revenues decline, government spending also falls. Due to the fall in government expenditure, GDP will tend to decrease. Whether GDP actually 1 8 falls or not will depend on the economy's response to lower priced imports and improvements in resource allocation as trade barriers are reduced. A decline in GDP need not imply a decline in real consumption, the latter often being the preferred measure of welfare. In the modeling exercise, GDP is estimated to fall in Kiribati, Palau, Samoa, Tonga and Vanuatu. However, all these countries experience increases in real consumption and employment. 1.38 Both Scollay et al (1998) and CGE (1998) observe the loss of tax revenue from trade liberalization and suggest that the PICs that do not already have a VAT system could adopt a broad-based tax to replace the taxes lost. In the modeling exercise, the introduction of such taxes would tend to reduce the estimated welfare gains.5 However, this does not mean that broad-based taxes should not be introduced. Given the actual distortions created by the present tax and tariff systems in may PICs, the efficiency gains from tax reforms would, in net terms, significantly enhance economic efficiency and growth prospects. The tax reform options available to PICs are discussed in chapter 2. 1.39 While the gains from liberalizing within a FTA with Australia and New Zealand are much larger than for a PIC-only FTA, Figure 1.5 shows that the gains for the PICs are greatest when there is complete MFN liberalization by each PIC (Scollay et al 1998 and Scollay 2001). But as CIE (1998) notes, the larger and the more comprehensive the liberalization, and the greater the welfare gains, the larger the structural adjustment that has to be made. Figure 1.5: Welfare Gains to PICs from Full MFN Liberalization (in % GDP) 8 7 - FTA with Aus. and NZ 6 Full MFN Liberalization 6 6 5 - 4 3- 2- 0 Source: Scollay (2001) and CIE (1998). 5This is because in the model, the comparison point is one where there are no VAT-related distortions. 19 Features of PICTA .0 The main terms of the PICTA, as negotiated, are as follows: According to Article 7, tariffs are to be progressively and automatically reduced according to graduated schedules. The schedules vary according to the initial level of tariffs, but all tariffs are to be reduced to zero by the "developing" PICs no later than 2010, and by the "small island states" and "least developed countries" no later than 2012 (see chapter 2 for details).6 * According to Article 9, non-tariff measures are to be eliminated upon entry into force of PICTA, or to be converted to tariffs within six months with advice to other parties. * Article 8 provides for a "negative list" of "Excepted Imports" to be listed in Annex In to the Agreement. Those countries which have not already provided their negative list must do so on or before ratification and accession. Annex IV sets out the schedule for elimination of tariffs on the excepted imports, which provides for all tariffs to be eliminated by 2016. Alcohol and tobacco products are exempted from liberalization for the first two years. of the Agreement (Paragraph 13, Article 8) during which time a study is to be undertaken on the integration of these important sources of tariff revenue into the PICTA. * Rules of origin will be based on a 40 percent value added criterion, with provision for cumulation of value added between members (Article 5 and Annex I). * Article 11 (Safeguard Measures), Article 12 (Dumped or Subsidised Imports), and. Article 13 (Balance of Payments) allow emergency action in the case of serious injury or the threat of serious injury to domestic industries or serious decline or threat of serious decline in a member's monetary reserves, but only as a last resort (Article 10). Prior to any such actions being taken, members are expected to carry out a public investigation at which all affected parties may make submissions. Any action taken should be in the "public interest" * Article 14 allows for tariff protection of "developing industries" (or "infant" industries) for a period of ten years, or 15 years in the case of measures taken by Small Island States or Least Developed Countries. * Government procurement is to be liberalized "as soon as possible" (Article 15). * Disputes between members should preterably be settled by consultation, but failing that by mediation or ultimately by arbitration (Article 22). Failure by the defaulting 6 PICs designated as small island states are Cook Islands, Kiribati, Nauru, Niue, Republic of Marshall Islands, and Tuvalu. PICs designated by the United Nations as least developed countries are Kiribati, Samoa, Solomon Islands, Tuvalu and Vanuatu. This means that only five countries-Federated States of Micronesia, Fiji, Palau, Papua New Guinea and Tonga-are required to eliminate their trade barriers by 2010. Of these, Federated States of Micronesia and Palau already have tariff levels close to zero. 20 member to comply with an arbitrator's award can be sanctioned by suspension of "concessions or the performance of any other obligations under the Agreement" by the affected member. * The Agreement encourages transparency between members with regard to any action covered by the Agreement, and for the regular review of any actions taken. It also provides for a general review of the Agreement each five years. 1.41 Scollay (2001a) notes that PICTA was designed to minimize negotiating and administrative costs, especially through the automaticity built into the tariff reduction schedules and the adoption of a "negative list" approach to exceptions. Also, while it allows for the full range of "emergency actions" to provide flexibility in responding to economic disturbances and to implement development initiatives, there is discipline in the application of these measures and encouragement of transparency between members at all times. 1.42 PICTA has many desirable features.- Although the tariff reductions are to take place over a long period, the commitment to agreed, phased reduction of tariffs is important to prevent "backsliding" on trade liberalization. Non-tariff barriers will be converted to tariffs and would be subject to the agreed phasing-out. Another very important feature of the agreement is that members will establish a mechanism for a public airing of the arguments for and against the maintenance of existing trade barriers or for the imposition of new trade barriers in the form of "emergency assistance". Such a mechanism has proven very useful in Australia to educate the public about the benefits and costs of trade barriers.7 This, of course, raises an issue regarding the appropriate authority to adjudicate on such trade-related matters. For example, a formal body along the lines of the Productivity Commission in Australia may be too costly for individual countries to undertake. Hence, a Forum-wide institution might be contemplated. A regional body would also maximize the objectivity of its deliberations-an important matter as small countries have difficulties in maintaining objectivity in the face of powerful vested interests. Being a Forum-wide body, it would also provide greater discipline on the use of protective measures than single-country institutions. A possible trade-off is that this would entail creating another regional body with its own bureaucracy. 1.43 A supporting program of trade and investment facilitation measures is also envisaged along with PICTA. There would be substantial benefits from the implementation of such a program. The main efforts would be directed at upgrading quarantine facilities, improved and harmonized customs procedures, standards and conformance enforcement, and liberalized government procurement. The implementation 7 The Australian Productivity Commission (formerly the Industries Assistance Commission and the Industries Commission) provides a venue in which all parties with an interest in government assistance to industries (importers, exporters, domestic producers and consumers) can express their views about the retention or imposition of forms of assistance to industries. 21 of these measures could lead to a large reduction in the transactions costs of businesses in the region.8 1.44 The effectiveness of PICTA should also be judged in terms of its effects in facilitating the widening of liberalization to other countries. A major concern, therefore, should be whether it creates or embeds interests that will resist widening of the liberalization. A concern about a comprehensive RTA such as PICTA with its protection of "infant industries" is that the trade diversion impacts will make it more difficult to move towards further liberalization. A new industry established under temporary protection that may be profitable within PICTA could be unprofitable as soon as there is a widening of liberalization to include other countries/regions. Besides the fact that the "infant industry" will become a new interest arguing against trade liberalization, the establishment of industries that will become uneconomic as soon as further liberalization occurs is wasting resources. 1.45 To avoid these pitfalls, members of PICTA that have high trade barriers against the rest of the World would be well advised to lower them. However, these are also the countries in which tariff revenues are more important. Therefore, for these countries, finding alternative government revenue sources is a high priority. 1.46 Article 14 potentially allows for the establishment of uneconomic activities behind protective barriers. However, governments are normally not in a position to 'pick winners' that would justify the start-up costs to the economy. Hence, it would be best for member countries to subject decisions about the use of Article 14 to the same form of public inquiry and public interest test as set down for application of Articles 11, 12 and 13. Again, a Forum-wide body would be more effective than individual national institutions. 1.47 In the Caribbean region, the integration of small states has focused on trade in goods. But goods trade is becoming a progressively smaller part of the total trade of these countries. Services production and trade is by far the most important activity of these countries, averaging 60 percent of GNP. Tourism is the largest services industry, but exports of financial, professional and business services are growing as well. The same is becoming true of the PICs. PICs have very low capacity to absorb industrial waste and many have poor agricultural and other natural resources, except for the natural beauty of their beaches and coral reefs. Hence, they are physically unable to develop much in the way of primary and secondary industries and will be forced to concentrate on services sector activities similar to those in the Caribbean. For this reason it will be 8 Poor quality is obviously constraining PIC exports. For example, Pacific Outlook, of March-April, 2002, (the publication of the Pacific Island Trade and Investment Commission-an agency of the Forum Secretariat) reported that a major Australian importer of Pacific Island fruits and vegetables had stopped imports from the region because of poor quality, due to poor post-harvest practices, poor packaging, and non-compliance with quarantine requirements. 22 important for PICTA to broaden its scope beyond merchandise trade and to move to deeper integration to allow free intra-regional movement of capital and labor.9 5. Moving beyond PICTA 1.48 Under PACER, the framework agreement, negotiations over a FTA between the PICs and Australia and New Zealand are to commence eight years after PICTA comes into force. In practice, negotiations might start earlier by mutual consent or triggered by the PICs commencing negotiations over iFTAs with other developed countries. When, for example, the PICs enter into negotiations over reciprocal free trade agreements with the EU-as envisaged under the Cotonou Agreementl°-this will presumably trigger negotiations with Australia and New Zealand. The trade provisions of the Cotonou Agreement provide for negotiation of reciprocal FTAs between ACP states and the EU to commence in 2002 and to conclude by 2007, with the existing non-reciprocal arrangements to continue in the meantime.'1 The Cotonou Agreement provides that where ACP states extend preferences to other developed countries that are more favorable than those applying to the EU, the same preferences must be extended to the EU. Thus, entering into negotiations over reciprocal FTAs with the EU-envisaged as being incorporated in Regional Economic Partnership Agreements (REPAs)-would mean beginning negotiations with Australia and New Zealand. Conclusion of a reciprocal PIC-FIA with Australia and New Zealand would initiate negotiations over a similar arrangement with the EU. 1.49 In addition, the three US Compact countries also have a provision in the Compact of Free Association with the United States whereby they must not give more favorable tariff treatment to countries other than the United States. Hence, reciprocal trade agreements negotiated with the EU or with Australia and New Zealand would trigger discussions with the United States over the extension of similar arrangements to the United States. 1.50 In sum, PACER and the Cotonou Agreement have set in motion a process of negotiation of FTAs between the PICs and the EU and Australia and New Zealand, and for providing the United States with similar preferential treatment. The widening of preferential trading arrangements beyond PICTA is inevitable. Only the timing, extent and benefits are uncertain. Are these steps towards widening trade liberalization the best possible for the PICs or is it preferable to take some other route? 9 Other forms of deeper integration include adoption of a single currency. Duncan and Xu (2000) and de Brouwer (2000) have argued that the investment environment would be greatly improved in those Pacific island countries presently maintaining their own currencies if they were to move to the Australian dollar. This would also facilitate integration between them. '1 Previously, the Lome Agreement, which was negotiated between the EU and African, Caribbean and Pacific (ACP) states that were former colonies and former or existing territories of EU countries (see Box 1.4). Fiji, Kiribati, Papua New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu and Vanuatu were members of the Lom6 Agreement. The Contonou Agreement has extended membership of the ACP Group to the other (six) PICs. " The Fourth WTO Ministerial at Doha agreed to grant the EU a waiver from its non-discrimination obligations while the FIAs with the PICs are being negotiated. 23 5.1 PICs and the EU 1.51 The EU has initiated the negotiation of Regional Economic Partnership Agreements with the ACP countries in order to make its preferential trading arrangements with these countries WTO-compatible. Negotiations over reciprocal trade agreements between the PICs-individually or in groups-and the EU are to commence in 2002 and to conclude by 2007 (see Box 1.4). 1.52 In the past, the countries that have taken advantage of preferential access to the EU market are Fiji, Solomon Islands and PNG. All three have benefited from the 24 percent nominal margin of trade preference on canned tuna. Fiji has also gained from the Sugar Protocol which has allowed it to sell its key export in the EU market at well above world market price. EU "sugar transfers" have amounted to 3.5-5 percent of Fiji's GDP in recent years. The history of lower-cost imports from the ACP Group suggests that industries such as sugar may still be treated as an exception even in a FTA with the EU. PICs that are included in the UN definition of least developed countries (Kiribati, Samoa, Solomon Islands, and Vanuatu) would benefit from the Everything But Arms (EBA) proposal of the EU that would allow duty free and quota free access to its market. For these countries there is thus little if anything to gain from forming a FTA in goods with the EU. The US compact countries have virtually no trading relationship with the EU. Thus for most countries there may not be much additional benefit by forming a FTA with the EU. Box 1.4: The Cotonou Agreement: ACP and EU Partnerships After Lome Given doubts about whether the WTO would extend its waiver with respect to the Lome Convention, the EU in a 1996 Green Paper on EU relations with ACP countries, examined various options for future ACP-EU relations. In the area of trade, the main option canvassed was a move to reciprocity and differentiation. The EU would no longer provide non-reciprocal preferences but would form PTAs with groups of ACP countries. The ACP countries would first form regional agreements among themselves and the EU would then form PTAs with them. However, the EU would treat least developed countries differently from the others by not requiring reciprocity. Following four Lome Conventions, the EU and the ACP countries agreed on a new relationship in the form of the Cotonou Agreement, signed in Cotonou, Benin in June 2000. The Cotonou Agreement between 15 European states and 77 ACP countries, proposes that ACP countries enter into Economic Partnership Agreements (EPAs) with the EU, either individually or on a regional basis (regional EPAs or REPAs). Negotiation on EPAs or REPAs are due to begin after September 2002 and will enter into force as of 1 January 2008. The new arrangements, which are to be in conformity with WTO provisions, are to involve progressive removal of barriers to trade between EU and ACP member countries. In the interim, the existing non-reciprocal access arrangements for the ACP countries will be maintained. For example, the existing sugar and beef protocols will be provisionally maintained, but will be re- examined within the context of the negotiation of the new trade agreements. After the new trade agreements enter into force, there will be a transitional period of at least 12 years. The new agreements will include cooperation in areas linked with trade, such as competition policy, the 24 protection of intellectual property rights, standardization and certification, sanitary and phytosanitary measures, trade and the environment, trade and labor standards, and consumer policy. Regional cooperation is considered to be the foundation for the new ACP-EU agreement. Regional integration is considered to be the foundation for the new trade regimes to be established. Article 35 of the Agreement states that "economic and trade cooperation shall build on regional integration initiatives of ACP states, bearing in mind that regional integration is a key instrument for the integration of ACP countries into the world economy". Regional cooperation is also encouraged in terms of macroeconomic policies. This follows the approach adopted within the EU of taking into account the regional dimensions in macroeconomic reform programs. Regional cooperation is also seen as a key development strategy in two respects. First, there is the regional integration through the removal of trade barriers. Second, there is the so-called "functional regional cooperation" (Article 30) which covers support to deal with common or cross-border problems-that arise in areas such as infrastructure, environment, education, and health. Regional integration is seen as a key factor in poverty reduction in the ACP countries through its enhancement of economic growth. It is also seen as providing a framework for the adoption of sound economic policies through regional surveillance and peer pressure. It is implicit in the Agreement that regional integration will only be effective if member countries adopt sound macroeconomic management. The important role of the private sector in achieving the benefits of regional integration is also recognized. Instruments to support private sector development are given a key role in the Agreement with the inclusion of an Investment Facility. Risk capital for private sector development will be provided through this Facility. The intended negotiation of WTO-compatible economic partnerships recognizes the benefits of regional cooperation within the three ACP regions. The Cotonou Agreement essentially recognized the valuable work that the Pacific Forum had done through promoting the external economic relations of its members by admitting the six remaining Forum Island Countries as new members of the ACP Group.* For programnming and project identification purposes, the Regional Authorizing Officer is the Secretary General of the Forum Secretariat. The accession of the six new members means that it will now be possible to prepare projects for all Forum Island Countries, rather than for only part of the group. Similarly, implementation of regional projects will be able to utilize existing Forum-wide regional institutions, rather than creating separate EU- Pacific ACP arrangements. * The new members are Cook Islands, Federated States of Micronesia, Republic of the Marshall Islands, Nauru, Niue, and Palau. 1.53 This so-called "hub-and-spoke" RTA would be more favorable to the EU (the hub) than to the other members (the spokes) for two reasons. First, there is the transfer of tariff revenue involved when tariffs are lowered. As tariffs in the PICs will often be higher than EU tariffs, the revenue transfer is likely to favor the EU. Second, investment is more likely to locate in the hub country than- in the spokes. By locating in the hub investors.can gain duty-free access to all countries with which the EU, has RTAs. As explained above, differences between the PICs with respect to their relationship with the 25 EU makes it difficult to have a single REPA with all of them. This could lead to separate REPAs with different groups of PICs or even EPAs with individual countries. It would also mean proliferation of the hub-and-spoke relationship. The concept of the EU forming REPAs with groups of ACP countries in different regions (Africa, Caribbean, and the Pacific) already creates incentives for investment to concentrate in the EU rather than locate in ACP countries. For these reasons it would be useful to explore whether there are benefits for the PICs in forming a partnership arrangement with the EU that covers areas where nearly all have an interest e.g. services like tourism, and the fisheries sector. 5.2 PICs, Australia and New Zealand 1.54 Australia and New Zealand are the source of a large share of PICs' imports. The formation of a RTA with Australia and New Zealand would lead to large losses in tariff revenue from those PICs with high tariff barriers such as Kiribati and Tonga. But this would not be the case where Australia and New Zealand is globally the cheapest source of imports-a situation likely to apply in only a few cases (because of lower transport costs or in the case of some primary commodities). 1.55 One of the main factors in the formation of FTAs and in the intensity of trade more generally is geographical proximity. The predominance of Australia and New Zealand in PICs trade can be seen in Tables 1.5 and 1.6. With respect to imports, Australia and New Zealand are by far the primary source, with the exception of the US Compact countries. 1.56 Exports are more widespread and generally relate to specific products in particular circumstances. At present nearly all exports from PICs enter Australia and New Zealand duty free under SPARTECA. The main exports are garments from Fiji and automotive wire harnesses from Samoa (see Box 1.5). Other exports are: sugar exports by Fiji to the EU and the US under preferential quotas; Solomon Islands' exports of canned tuna to the EU-again under preferential arrangements-and log exports to Japan; and Tongan exports of squash to Japan. Because of geographical proximity, closer integration with Australia and New Zealand is likely to have a greater pay-off in terms of trade expansion than integration with other regions-except perhaps for the East Asian region which offers complementarity in trading patterns with the PICs. Australia and New Zealand also offer considerable scope for deep integration along the lines of the member countries of the EU. 1.57 A potential gain from deeper integration with Australia and New Zealand could result from the freer movement of labor and capital. While investment can move without restrictions from the PICs to Australia and New Zealand, investment cannot move freely in the opposite direction. It is ironic that countries so short of capital hinder its entry through direct and indirect barriers. These barriers take the form of complex foreign investment legislation, with a number of restricted or reserved categories that are not open to foreign investors. Even where foreign investment legislation is relatively simple and liberal, administrative barriers in the form of lengthy approval mechanisms effectively discourage investment. 26 Table 1.5: PICs - Sources of Imports (% of total imports) Australia New Europe Japan USA Zealand Fiji 40.2 14.6 1.4 10.6 5.4 FSM 11.0 0.3 0.9 26.7 41.8 Kiribati 44.1 4.2 - 14.0 6.1 Marshall Is 2.6 1.0 - 4.3 61.7 Palau 1.4 0.3 - - 50.8 Samoa 26.0 33.7 2.0 7.9 12.9 Solomon Is 16.1 - 0.8 3.5 0.6 Tonga 26.5 35.4 0.0 5.3 14.5 Vanuatu 43.9 12.0 7.1 - PNG 55.0 4.8 2.6 6.5 3.7 Data are averages for the latest three years for which data are available. Data for FSM and PNG, are derived from partner country sources in the UN COMTRADE database accessed through WITS software, and refer to 1999. Country data sources are used for all others. Table 1.6: PICs - Destination of Exports (% of total exports)* Australia ZNew Europe Japan USA Zealand Fiji 37.0 4.9 19.3 16.9 4.3 Kiribati 3.3 0.0 6.7 5.4 13.5 Samoa** 4.2 9.3 15.4 0.6 17.7 Solomon Is 1.7 - 9.9 40.9 1.0 Tonga 3.3 11.2 0.0 49.5 15.1 Vanuatu 4.0 2.1 33.1 14.3 *Country data sources have been used. Data are averages for the latest three years for which data are available. For countries for which data are not available, UN COMTRADE was tried, but partner country data are unreliable e.g. fish caught by distant water fishing nation fleets are included as exports. **The main destination is American Samoa accounting for 50 percent of exports. Yazaki's exports of automotive wire harnesses, a key manufactured export, to Australia are not reported as exports by the Samoan authorities. 27 1.58 Labor can move relatively freely from several of the PICs (e.g. Samoa, Tonga, and FSM) to metropolitan countries such as Australia, Canada, New Zealand, and the United States. For the Melanesian groups of countries (Fiji, Papua New Guinea, Solomon Islands and Vanuatu) such labor movement and emigration is not available. Labor movement in the opposite direction-such as from Australia and New Zealand-is generally tightly controlled throughout the PICs. There appears to be a widespread view in the PICs that employment of labor from high-income countries is a negative-sum game, with such labor displacing local labor. But as Duncan and Lawson (1997) found in their study of business costs in Papua New Guinea, expatriate labor and local labor are complementary-employment of expatriate labor means increased employment of local labor. 1.59 Extending PICTA or the FTA with Australia and New Zealand to cover services is likely to increase its benefits as the experience from the Caribbean shows. Air transport and commercial shipping appear to be two obvious examples for the PICs. One aspect of service liberalization is that service providers may need to be established locally and this may involve labor mobility to establish temporary or even permanent residence. It would be worthwhile to begin a debate on the issue of labor mobility among the PICs, Australia and New Zealand. Box 1.5: SPARTECA - PICs Preferential Access to Australian and New Zealand Markets, and its Erosion The South Pacific Regional Trade and Economic Cooperation Agreement (SPARTECA) is an agreement between Australia and New Zealand on the one hand and the island members of the Forum on the other, covering the preferential import of products by Australia and New Zealand. SPARTECA came into force on June 30, 1982. It is a non-reciprocal agreement whereby specified products originating from the PICs may enter Australia and New Zealand duty free and without quantitative restrictions. At present, nearly all exports from the PICs to Australia and New Zealand enter under SPARTECA. SPARTECA, when taken together with the provisions of Australian sectoral policy towards motor vehicles and textiles, clothing and footwear (TCF) industries, has been of substantial benefit to Samoa and Fiji. The Australian Motor Vehicle Plan which is a transitional arrangement for that sector specifies that in order to meet the definition of 'Australian-produced' there are obligations for domestic producers to meet certain domestic processing requirements. It was the change in the definition of domestic content in the early 1990's to include PIC-content that induced Yazaki (Australia) to shift its labor intensive production of automotive wire harness from Australia to Samoa. At its peak there were about 3,000 people employed at the facility but employment has now fallen to around 2,000. As Australia has moved to further liberalize its motor vehicle industry the domestic processing requirements have undergone changes in order to assure full compliance with the TRIMs agreement. This has eroded the preferential access granted to Yazaki. Similar concerns exist in Fiji over Australian liberalization of the TCF sector. The industry developed mainly as a 'cut, make arid trim' (CMNT) activity, and grew consistently from an export value of F$117 million in 1992 to F$333 million in 2000. By the end-1990s, 20,000 jobs had been created in the Fiji garment industry as New Zealand and then Australian manufacturers moved off-shore to capitalize on the benefits of the liberalization of the sector and its full 28 inclusion under SPARTECA from 1988. Australia, using an export subsidy regime for its textile sector, subsidised the exports of textiles which were then processed in Fiji, and other countries, and re-exported as garments to Australia. Generous tax holidays granted to manufacturers by the Fijian govemnment were an added incentive to locate in Fiji. However, as Australia moves to bring into effect the terms of the WTO's Agreement on Textiles and Clothing the Fiji garment industry will lose the benefit of subsidies provided to exported Australian textiles. Furthermore this will also compound their difficulties of complying with the 50% ex-factory cost rules of origin provisions of SPARTECA*. Thus changes in the trade regime in Australia which will assure WTO compliance would adversely effect both Samoa and Fiji. In Fiji a number of garment factories have closed recently. This is due to a number of factors -- the political and economic instability created by the 2000 coup, Australian and NZ trade union decisions to ban imports after the coup, end of the tax holiday period, but also importantly the uncertainty created by the changing trade regime. *At present Fiji garments are exported under a temporary derogation from the 50% rule for a range of products. This derogation is reviewed on an annual basis. Source: Grynberg (2000) and Bank staff consultations. 1.60 Based on other experience, the formation of RTAs with the EU or with Australia and New Zealand could be beneficial to the PICs, with benefits flowing in the form of technology transfer, cheaper inputs, and the locking-in of economic policy reforms along the lines suggested by Duncan (see Section 2.1). In order to benefit from improved market access that the RTAs would provide, it would be crucial for the developed country partners to provide technical and financial assistance to enhance supply-side capacity in the PICs. Trade-related capacity challenges confronting the PICs include meeting export market product standards, improving trade logistics and customs clearance, upgrading transport, communications and other basic infrastructure services, protecting intellectual property, and regulating service industries. Trade development and facilitation training will be very important in assisting potential exporters to move into international markets. 1.61 A key constraint to the development of export industries in liberalizing economies is that domestic firms have had little or no experience in exporting. Therefore, they have no established international distribution and marketing networks. A proven way of overcoming this hurdle is to establish joint ventures with foreign firms that already have established marketing and distribution networks. It will be important, therefore, for laws and regulations relating to foreign investment to provide an encouraging environment for such joint ventures to be established. Providing an encouraging environment for foreign investment to establish joint ventures will be even more crucial if trade liberalization is widened to include Australia and New Zealand, the EU and the United States, as exporting into these major markets will need such assistance even more. 29 6. The WTO Alternative 1.62 At present only Fiji, PNG, and Solomon Islands are members of the WTO. By liberalizing within the WTO framework, the PICs would gain improved market access to the rest of the world without transferring tariff revenue to the EU or Australia and New Zealand or to the US. The only way to achieve this in a RTA is by lowering tariffs to zero for all imports, which essentially amounts to unilateral trade liberalization. The WTO provides the lock-in benefits of an international treaty-probably to a greater extent than a RTA-because the WTO provides both rewards in terms of financial and technical assistance as well as sanctions. WTO membership also captures all the benefits of technology transfer from integration with other countries, without the problems of hub-and-spoke arrangements. 1.63 While the transaction costs of negotiating accession to the WTO are high, the negotiations to be entered into with the EU, Australia and New Zealand, and the United States would also be extensive and complicated. With prospective membership of multiple trading blocs, complexities arising from the need to reconcile the requirements of different blocs, the incompatibility of rules of origin, and incompatibility of legal and other institutions is almost unavoidable. Resolving these issues will place a huge burden on the limited skilled resources of the PICs.12 1.64 However, the recent experience of PICs with the WTO accession process has not been positive. Tonga applied to join the WTO in June 1995, Vanuatu applied in July 1995, and Samoa applied in April 1998. None of the three has yet acceded to the WTO. Vanuatu, after having spent over US$400,000 or over US$2 per capita recently broke off negotiations when its accession process was nearly complete. Lack of agreement with the US, with whom Vanuatu has very limited trade, was the issue on which the process faltered. US demands for opening up the retail sector and the telecommunications sector were viewed as too onerous for a least developed country . In the telecommunications sector France Telecom has a monopoly till 2012. Vanuatu believes, however, that opening up its telecommunications sector would lower the incentives for France Telecom to invest in necessary service improvements. Telecommunications infrastructure is vital for this remote country that is highly dependent on services and tourism. The retail sector has been traditionally reserved for ni-Vanuatu and is politically a very sensitive issue. It is feared that even a relatively small sized foreign investor could wipe out many small businesses. 1.65 Another pervasive problem for small economies within the WTO context is the high cost of maintaining membership, with the need for representation in Geneva to attend up to 50 meetings a year. Small countries could gain most of the benefits of WTO membership without having to join the WTO through unilateral trade liberalization. However, they would not have access to the financial and technical assistance available through the WTO to assist with adopting best practice institutions and regulations. Also, they would not have access to the protection provided by the WTO dispute resolution 12 Although it must be pointed out that bureaucrats have a vested interest in complexity as it ensures larger budgets and increased activity. 30 procedures.'3 Hence, consideration has to be given to making WTO membership accessible to small countries at less cost than at present. 1.66 "Behind the border" legal and institutional reforms in the PICs will be a key component in expanding trade, be it among themselves or with the rest of the world. These include reforms in areas such as sanitary and phytosanitary standards, technical standards, investment legislation, competition policy, and property rights. Enhancing supply-side capacity through technical assistance in designing and implementing these and other reforms in the special circumstances faced by PICs could be a critical role played by PACER and Cotonou. These would also help in the implementation of WTO commitments. 1.67 With the Cotonou Agreement in place and the PACER agreement signed with Australia and New Zealand, it is likely that PICs could face pressure from their high- income partners to become involved in negotiations over the envisaged RTAs. Trade liberalization within the WTO multilateral context could pre-empt the need for such negotiations. This would also have the benefit of freer trade with the Asian region-a region presently left out of the existing and proposed trading blocs, but a source of the world's cheapest imports, particularly in manufactures. 1.68 Ultimately, the negotiation of RTAs between the PICs and Australia and New Zealand and the EU will involve the exclusion or restriction of some of the exports in which the PICs have a comparative advantage. It is often the case that the so-called "preferential" trade agreements between the high-income countries and the developing countries allow free access for all commodities except those which the developing countries are best at producing. For example, under SPARTECA Australia allows free entry to all exports from Fiji, except for sugar which is banned. Similarly, the EU and the United States allow free entry to Fijian exports except for sugar which is brought in under quota. Beef exports from Vanuatu, its primary exports, are subject to EU quotas. It is unlikely that the international sugar market will be made any freer as a result of the next negotiating round of the WTO as domestic interests in the EU and the United States seeking protection of the sugar market remain as strong as ever. However, the chances of developing countries gaining freer market access for their exports are likely to be better in a WTO negotiation than in a bilateral negotiation over a RTA. 7. Conclusion 1.69 The PICs have chosen the formation of a free trade area amongst themselves as a first step towards fuller integration with the global economy. As a move that represents a shared view among PIC policy makers on the benefits of moving towards a more liberalized trade regime, PICTA is a positive step. However, for the 'stepping stone' not to become a 'stumbling block' the strategy has to be carefully managed: policies that 13 Bans on imports of kava by several Western European countries have adversely affected this new export market activity in countries such as Samoa and Vanuatu. If they were WTO members, the WTO regime for settlement of disputes over sanitary and phytosanitary issues would have provided a robust mechanism for the settlement of this dispute over the health effects of kava. The WTO agreement provides for the establishment of a scientific panel which would objectively assess evidence on kava's health impacts. 31 create interests which could stall the move towards wider liberalization need to be avoided. 1.70 The estimated economic gains from PICTA are very small. By the same token, it is likely that the structural adjustments involved will also be small. However, there could be trade diversion to the more industrially-advanced members, and redistribution of tariff revenues from importing members to firms in the exporting countries. In other cases of trading blocs among small developing countries, such trade and tariff diversion has given rise to tension and-the breakdown of agreements. 1.71 To minimize the problems of trade diversion and loss of tariff revenue-and the tension that this could create-the PICs are advised to lower their trade barriers against all countries. This will provide the greatest benefits from trade liberalization. But it will also involve larger losses of tariff revenue and greater structural adjustment. These are issues taken up in the following chapters. 1.72 PICTA has desirable features. These include the provision for public inquiries to establish the "public interest" in the retention of tariffs or non-tariff barriers, or in the case of "emergency assistance" with respect to the use of safeguards against sudden inflows of imports, dumping, and balance of payments problems. From Australian experience, this kind of public inquiry at which all interested parties can make representations is an excellent mechanism for promoting transparency and public education about government assistance to business activities. However, it is a process that is demanding of human resources and that requires impartiality. Hence, the possibility of devejoping a region-wide mechanism should be explored. 1.73 PICTA allows for government assistance through tariff protection for so-called "developing industries". Government has a legitimate interest in developing business activities. However, government assistance to special industries is discriminatory and, in particular, may damage export prospects in other sectors ("a tax on imports is a tax on exports"). Therefore, it would be useful to subject proposals of this kind to the same public interest test to be used in the case of emergency measures. The creation of uneconomic activities through government support has broader implications because firms being supported can become advocates for resistance to further trade liberalization. 1.74 The proposed formation of RTAs between the PICs and high-income countries- Australia and New Zealand and the EU-will generate much greater benefits for the PICs than PICTA, as modeling work on a RTA with Australia and New Zealand has shown. The increased benefits result from the fact that the high-income countries are usually sources of lower-cost imports than PIC members. As well, investment may be attracted to the PICs from the high-income partners because the lower-income countries are a source of lower-cost labor or posses natural resources. Through the cheaper imports and investment there would be a transfer of new technology and new knowledge. 1.75 However, counteracting these benefits will be trade and tariff revenue diversion to the high-income countries. There is also the "hub-and-spoke" problem with respect to investment in overlapping RTAs. An incentive is created for investment to concentrate in 32 the country that is common to the various trading arrangements -- the "hub" country, in this case the EU -- where it gains preferential access to all the "spoke" countries. 1.76 RTAs provide indirect economic benefits because they are a means by which members can lock-in their trade and other economic reforms. Commitment to an international obligation protects against pressures from-domestic interests to over-turn the reforms. But for the lock-in to be most effective in providing stability of policies (and thereby giving investors increased confidence), there should be both rewards and sanctions. It is unlikely that there will be a system of. rewards and sanctions in a RTA such as PICTA. In a wider bloc including high-income countries, rewards can be made available by the high-income country, but policy lock-in will be effective to the extent that sanctions would be applied. 1.77 Whether moving forward on PICTA or extending regional integration to Australia, New Zealand and the EU, greater benefits will flow from deeper integration. In particular, this should include trade facilitation measures and greater freedom of labor and capital. Such deeper integration within the Pacific region would also be beneficial if the focus for trade liberalization becomes the WTO. 1.78 Another option available to the PICs is the lowering of trade barriers on a multilateral basis in a WTO context whereby the PICs will gain the benefits of market access to the rest of the world in exchange for their barrier reductions. But PICs that have recently tried to pursue this route have found it a costly and uphill process. It is thus essential to explore how the current process of WTO accession as it affects them can be modified to reduce the problems and costs that they face. Developed country partners have a key role to play in easing the accession process through the provision of technical and financial assistance. While the transaction costs of negotiating WTO accession are high, they need to be viewed against the costs of negotiating multiple, overlapping RTAs. 33 2. INTERACTION OF TRADE LIBERALIZATION AND TAX REFORM 1. Introduction 2.1 Tariffs or other taxes on imports are an important revenue source in the region, particularly for the smaller island nations. The reduction in tariffs to take place under the Pacific Islands Countries Trade Agreement (PICTA) has the potential to erode this revenue. As trade between the PICs is small, the direct loss in revenue will be limited. However, the revenue loss can be expected to rise over time for at least some countries as trade diversion takes place. PICTA is envisaged as a first step to wider trade liberalization, in the first instance with Australia, New Zealand, and the EU. It was argued in chapter 1, that to maximize gains along this path, PICs should lower tariff barriers not only amongst themselves but with all trading partners. This strategy would lead to greater benefits for consumers, but may entail larger revenue loss. The good news is that alternative ways of raising revenues are available that could both offset revenue loss and make beneficial trade more likely under PICTA. This chapter examines some tax reform options that are available to the PICs to achieve these objectives. While the focus of this chapter is on raising revenue, the expenditure side is equally important for government policy. To the extent that government expenditures can be rationalized and reduced, the need to offset potential revenue loss would also be reduced.' 2.2 The chapter begins by outlining in Section 2 the broad features of the current tax and tariff systems in five selected PICs - Fiji, Kiribati, Samoa, Tonga and Vanuatu. It shows that reliance on taxes on international trade is high especially in countries without a VAT. In Section 3 estimates of potential revenue loss from PICTA based on earlier studies are provided. Section 4 considers in detail the experience of the selected countries that have adopted the VAT - Fiji, Samoa, and Vanuatu. Having a broad-based consumption tax like the VAT implies that they have already implemented a key reform suggested for countries undertaking tariff reductions. Adoption of the VAT has helped make up for revenue losses arising from tariff reductions that have taken place in the past. The implementation of the VAT has, however, not been without problems. Some that arose in this context relate to zero rating, exemptions, thresholds, compliance and administrative capacity. The chapter then analyzes in Section 5 the impact that PICTA would have on these countries. In general the impact on revenues is small and can be managed with modest reforms. But given that greater benefits are likely to be had by liberalizing beyond PICTA, some directions for the larger reform effort that this entails are provided. In all countries a reform of the excise system would be beneficial and relatively easy to implement. Other reforms include, taxation of public enterprises, removing the exemptions on export income, and bringing large scale primary production ' Improving public expenditure management in PICs was examined in detail in World Bank (1998). The analysis showed that there was considerable scope for streamlining and reducing government expenditures. Many issues discussed there remain relevant, especially the need for governments to focus on core functions e.g., providing a stable macro environment; improving the institutional infrastructure such as property rights, law and order; providing basic education, healthcare and economically viable physical infrastructure; and protecting the environment. 34 in to the tax net. Section 6 considers the situation of Kiribati and Tonga. These two PICs are more exposed to revenue loss and do not yet have a broad based consumption tax. The potential impact of PICTA on these countries is analyzed by identifying products that they currently import, but could be provided by other PICs if they have preferential access. In line with earlier studies it is found that Kiribati and Tonga may well lose 20 percent and 15 percent respectively of their tariff revenues. However, these losses can be made up by moving to a broad-based consumption tax. The type of tax that these countries adopt must be tailored to their relatively simple production structures, and administrative capacity. There are lessons from the experience of other PICs who have adopted a VAT. The key trade-off is between the efficiency gains from a VAT-like tax in a very small and undiversified economy versus administrative costs. Section 7 is a concluding section that summarizes the main findings and lessons. 2. The Structure of Revenue in Selected Pacific Island Countries 2.3 This chapter is focused on the position and experience of five PICs-Fiji, Kiribati, Samoa, Tonga and Vanuatu-selected to capture the diversity of economic environments seen amongst the PICs. 'Along with Papua New Guinea, Fiji is the most industrialized of the PICs, offering relatively well-developed agriculture, tourism, processing and service sectors. The economy of Kiribati is representative of the island atoll nations of the region with a high govemment presence, a very small private sector and almost no processing activities or prospects for such activities. Vanuatu, Samoa and Tonga are medium-size PICs in terms of population, have relatively small processing sectors, but benefit from a good resource base for tourism, agriculture and fishing. Fiji, Samoa and Vanuatu have pursued comprehensive economic reform programs for a number of years including tax and tariff reform, with varying levels of success. In Tonga a comprehensive public sector and tax reform program is about to be launched. Table 2.1: Structure of Government Revenue for Fiji, Samoa, Tonga, Vanuatu and Kiribati Fiji Kiribati Samoa Tonga Vanuatu 2001 2001 2000/01 2000/01 2001 % of total revenue and grants Tax revenue 82 24 64 70 84 Income tax 28 8 14 13 - Tax on goods and services 35 - 37 11 48 VAT or other sales tax 28 - 18 8 35 Excise duties 7 - 19 - - Other - - - 3 13 Tax on international trade 16 15 12 46 34 Other tax - - 1 - 2 Non-tax receipts 17 48 10 27 10 Grants - 28 26 3 6 Total rev & grants (US$, mil) 400 46 84 40 48 Tax revenue as % of GDP 19 28 23 20 19 35 Figure 2.1: Trade Taxes as a Share of Tax Revenue* 70% 60% 50% 40%/ 30% - 20%- 10%/ 0% Fiji Kiribati Samoa Tonga Vanuatu *Latest year for which data are available 2.4 The structure of government revenue for these five PICs is shown in Table 2.1 and Figure 2.1. There is considerable variation in the contribution of the different taxes in these five PICs. Fiji, Samoa and Vanuatu have VAT-type taxes that are making important revenue contributions and were introduced with a substantial reduction in tariffs. The VAT rate is 10 percent in Samoa2 and Fiji, and 12.5 percent in Vanuatu. Reflecting its relatively high level of per capita income, Fiji collects a much larger share of its revenue from income taxes, and has a relatively diverse revenue structure. Samoa is the least reliant on trade taxes and collects a relatively high level of revenue from excise duties. Kiribati and Tonga are highly reliant on trade taxes. Tariff collection rates (the ratio of tariff revenue to the value of imports) and actual tariff rates are presented in Figure 2.2 and Table 2.2 respectively. Tariff collection rates are typically much lower than what would be expected on the basis of published tariff schedules due to widespread duty exemptions and under-collection. In Kiribati, non-tax revenue (especially fishing license fees) is the most important revenue source. Non-tax receipts (other than grants) are also quite substantial in Fiji and Tonga. Grants are important in Kiribati, Samoa and Vanuatu. Vanuatu is the only PIC where there are no direct taxes on income. 3. Impact of PICTA on Government Revenue - Estimates from Earlier Studies 2.5 An attempt to estimate the potential revenue loss of a trade agreement among the PICs was undertaken by Scollay et al (1998) using a general equilibrium approach. Some key results are summarized in Table 2.3. The modeling estimated significant net losses in tariff revenue for the Cook Islands, Kiribati, Niue, Samoa, Tonga, Tuvalu and Vanuatu. The percentage losses in tariff revenue tended to correspond to countries with a combination of higher than average tariff levels and/or a relatively large share of imports sourced from other PICs. The lower impact in Fiji together with the positive impact on imports from the rest of the world suggests that Fiji gains from trade diversion effects, i.e. 2 This will be raised to 12.5 percent later this year. 36 it is likely to increase its exports to other PICs. Papua New Guinea experiences a very large percentage increase in imports from PICs but from an extremely small base (as specified in the model's database). Figure 2.2: Tariff CoUection Rate* 35% 30% 25% - 20% - 15% 10% , 5%- 0% Fiji Kiribab Samoa Tonga Vanuatu *Average for 1998-2000. Table 2.2: Tariff Rates by Broad Categories (%) Fii Kiribati Samoa Vanuatu Tonga** Prepared, fresh, hilled meat & fish 3 - 27 23 - 78 0 - 10 10 - 55 15 - 25 ther Basic foods* 0 - 27 3 - 43 0 - 20 0 - 30 15 & 25 Most other onsumption goods 0 - 27 3 - 63 10 - 20 0 - 25 0 - 20 Tobacco & alcohol high specific high specific high specific rates 3 - 235 rates 15 - 125 rates lect. & mech. appliances 0 - 10 3 - 53 10 - 20 0 - 30 15 Petroleum,*** metals nd minerals 3 -10 3 - 38 10 - 20 0 - 25 0 - 20 *Vegetables, fruit, cereal, flour, bread, pasta, eggs, dairy, coffee, spices ** There is in addition a 20% ports and services tax which in effect is an import duty. High specific rates apply to some petroleum products 37 Table 2.3: Impacts on Imports and Tariff Revenue of a PIC-FTA (based on pre-1998 data) Tariffs as a Imports Percentage change In Imports % net change % of from PICs From PICs From rest of in tariff Imports as % total as a result of world as a revenue as a Imports FTA result of FTA result of FTA Cook Islands 11 10.8 7.9 -0.5 -11.2 FSM 1 0.0 0.1 0.0 -0.0 Fiji 14 0.4 7.7 0.2 -0.4 Kiribati 28 20 13.3 -2.6 -21.9 Marshall Is. 9 1.0 10.0 0.1 -0.9 Nauru 0 3.6 0.1 0.0 -3.6 Niue 9 9.0 6.6 -0.1 -9.1 Palau 5 0.0 3.4 -0.0 -0.2 PNG 14 0.7 92.5 -0.3 -0.7 Samoa 22 14.0 12.0 -1.0 -15.0 Solomon Is. 12 3.5 6.1 0.2 -3.4 Tonga 18 7.2 10.5 -0.1 -7.4 Tuvalu 13 21.1 3.1 0.3 -21.4 Vanuatu 29 9.3 2.7 2.6 -28.5 Source: Scollay et al (1998). 2.6 In presenting his findings, Scollay et al (1998) noted the various limitations of the data and the underlying economic model (pp.56-58). It is important to recognize the highly aggregated nature of his model with just four sectors for most countries and six sectors for Fiji and Papua New Guinea. Such a level of aggregation is unlikely to reliably capture the extent of trade diversion for specific commodities since aggregated elasticities and trade data will tend to underestimate trade diversion effects. These results should be interpreted as indicative only of the changes that could be expected, and not the changes that will actually occur. Given the limitations of the data, emphasis should be placed on the relativities rather than the size of the numbers, and whether the numbers are positive or negative. 2.7 In a follow-up study, Filmer and Lawson (1999) argued that Scollay et al overestimated the potential tariff revenue loss because the database included in PIC trade substantial imports that were consigned through Fiji but did not originate there. In addition they noted that since the Scollay et al (1998) study was undertaken, the tariff collection rate (the ratio of tariff revenue to the value of imports) had fallen substantially in the Cook Islands, Samoa and to a lesser extent Vanuatu (see chapter 1). For example, by 1999 the tariff collection rate in Samoa had fallen to less than half that which applied to the Scollay et al analysis. This means that if the Scollay et al estimates were updated using refined data but the same model structure, the revenue loss would be lower than originally estimated. 38 2.8 Filmer and Lawson presented altemative estimates of the potential revenue loss taking account of more recent changes in tariff rates. They concluded that only Kiribati and Tuvalu would experience significant direct revenue losses from PICTA, amounting to 13 to 14 percent of total tax revenue and probably considerably less if allowance was made for non-PIC imports sourced through Fiji. Tonga was projected to experience a loss of a maximum of 6 percent of total revenues and other countries were projected to experience a loss in the range of 0 to 4 percent of total revenues. A comparison of the Scollay et al and Filmer and Lawson estimates are provided in Figure 2.3. 2.9 The Filmer and Lawson estimates are based on a static approach that assumes that PICTA has no impact on the level of trade. Under this approach, the revenue loss is assumed to be limited to the tariff revenue currently collected on trade with other PICs. This means that their estimates do not take into account the potential for PICTA to lead to trade diversion. Trade diversion would see duty free imports from other PICs displace non-PIC products, and lead to further revenue losses. Figure 2.3: Alternative Estimates of the Potential Loss in Revenue of a PIC-FTA (pre-1998 data) FSM Nauru D Filmer & Lawson (static estimate) Palau Fiji Scollay (general equilibruim estimate) PNG Marshall Islands Solomon Islands Tonga Cook Islands Niue Samoa __ T uvalu Vanautu Kiribati 0 3 6 9 12 15 Net loss in tariff revenue as a per cent of tax revenue 39 2.10 Although the potential extent of trade diversion is somewhat unclear at this stage, it seems likely that the main adverse impacts under PICTA would occur for Kiribati, and to a lesser extent Tonga and Vanuatu. If more general trade liberalization was undertaken all PICs would face a much more significant impact on their revenue and substantial tax reform would ultimately be required in several PICs. These issues are discussed sections 4, 5 and 6 below. For two of the more exposed economies, Kiribati and Tonga, the approach adopted to estimate the potential revenue loss identifies goods that have relatively high tariffs, that are produced in other PICs to a substantial extent, and that are considered likely to be candidates for trade diversion. 3.1 The Timing of Revenue Impacts 2.11 A key factor affecting the potential impact of PICTA will be the willingness of countries involved in a preferential trade agreement to expose their local producers to competition. If member countries are reluctant to do so, and make wide use of the option provided under PICTA to declare locally produced goods threatened by potential competition as exempt from PICTA, little trade and hence little revenue loss would occur. However, as noted in chapter 1, such a strategy would be sub-optimal for the PICs as it would limit the benefits of trade liberalization. 2.12 In considering the potential revenue impact, the pace at which the new trade agreements are phased in needs to be kept in mind. For example, under PICTA the Least Developed Members and Small Island States are to remove tariffs on originating goods by no later than 2012 while the deadline is 2010 for larger countries (Table 2.4). Member countries can also declare a negative list of exempted imports. The Agreement sets a maximum tariff on exempted imports of 50 percent in 2007 for Least Developed Members and Small Island States and 40 percent for other members, but these tariffs must be removed gradually over the period to 2016. This phased approach to implementing PICTA means that a member may choose to incur only small revenue losses over the short term and delay incurring greater losses until well into the future. Some members may instead choose more rapid tariff reductions accompanied by tax reforms to regain the lost revenue. 2.13 PACER provides for the negotiation of a free trade agreement that includes Australia and New Zealand no later than eight years after commencement of PICTA. These negotiations could take place earlier if the PICs extend preferential trade access to a non-PIC (e.g. the European Union, as provided for under the Cotonou Agreement) or as otherwise agreed. Australia and New Zealand account for a large share of most PICs' imports (see chapter 1). Given the diversity of these economies, the provision of a significant price advantage over competing countries via tariff concessions could be expected to result in a substantial increase in the share of PIC imports accounted for by Australia and New Zealand. Thus PICTA may initiate the erosion over the long term of a large proportion of tariff revenue collected by the PICs. The speed at which tariff revenue is lost would depend on the nature of the preferential trade agreements put in place. 40 Table 2.4: Schedule for Tariff Reductions under PICTA Ad valorem tariffs: Fiji FSM, PNG, Tonga Maximum tariff on goods from: Base tariff on goods on the Entry into force 1.1.2004 1.1.2006 1.1.2008 1.1.2010 entry into force of this of this Agreement Agreement More than 20% 20% 15% 10% 5% 0% More than 15% but not 15.% 10% 5% 0% more than 20% More than 10% but not more than 15% 10% 5% 0% Not more than 10% 0% Small Island States and Least Developed Countries: Cook Is., Kiribati, Marshall Is., Nauru, Niue, Solomon Is., Samoa, Tuvalu, Vanuatu Maximum tariff on goods from: Base tariff on goods on the 1.1.2004 1.1.2006 1.1.2008 1.1.2010 1.1.2012 entry into force of this Agreement More than 25% 25% 17.5% 10% 5% 0% More than 20% but not 20% 15% 10% 5% 0% more than 25% 20 _______ More than 15% but not 15% 10% 5% 0% more than 20% More than 10% but not 10% 5% 0% more than 15% Not more than 10% 5% 0% Specific and Fixed Tariffs: Fiji, FSM, PNG, Tonga Maximum specific or fixed tariff on goods, as a perce tage of base tariff (value), from: Entry into force of 1.1.2004 1.1.2006 1.1.2008 1.1. 2010 B this Agreement % of Base | 80% 60% 40% 20% 0% Tariff (value) 8%4%2%0 Cook Is., Kiribati, Marshall Is., Nauru, Niue, Solomon Is., Samoa, Tuvalu, Vanuatu Maximum specific or fixed tariff on goods, as a percentage of base tariff (value), from: 1.1.2004 1.1.2006 1.1.2008 1.1.2010 1.1.2012 % of Base 80% 60% 40% 20% 0% Tariff (value) Source: Forum Secretariat (2001) 41 4. Experience with a VAT in the Region 2.14 A number of countries in the Pacific region have introduced VATs as part of tax and tariff reform programs, and this experience provides important lessons for the other PICs. The PICs operating a VAT-type tax are Samoa, Fiji, Vanuatu, the Cook Islands and Papua New Guinea. The VAT is attractive from an economic efficiency perspective, mainly because it reduces the cascading of taxes. Cascading occurs when taxes are imposed on taxes paid at an earlier stage of production or distribution. A VAT reduces this cascading because it provides a tax credit on most if not all taxes paid on inputs. Since taxation of inputs tends to distort production decisions their removal would make an economy more efficient. A VAT has the added advantage of improving the record keeping of business. 2.15 In this section the experience of Fiji, Samoa, and Vanuatu with the VAT is briefly reviewed. Having a broad-based consumption tax in place implies that they have already implemented a key reform suggested for countries undertaking tariff reductions. However, given that these countries still continue to rely on trade taxes to a significant extent, other forms of raising revenue may need to be explored as trade is further liberalized. 2.16 In general, the VAT has helped Fiji and Samoa make up for revenue loss from tariff reduction. In Vanuatu, where the VAT is more recent there have been revenue shortfalls. Each country experience offers lessons for other countries in the region that are contemplating moving to broad based consumption taxes. Some of the problems faced in the context of the VAT relate to zero rating, exemptions, thresholds for compulsory registration, refunds, audits, and collection costs. 4.1 Trends in Taiff and VAT revenue 2.17 Fiji has undertaken a number of important tariff and tax reforms since 1989. In the 1989 budget, the govemment commenced a major trade liberalization program replacing import licenses with tariffs and subsequently reducing average tariffs significantly. The standard fiscal tariff fell from 50 percent in 1989 to 25 percent in 1993, and further to 15 percent in 1997. 2.18 After 1993, tariff reform slowed and to some extent was characterized by a stop- start process. For example, in 1998, duties on imported items produced locally were raised to 35 percent with a promise to reduce the rate progressively each year. In the 1999 budget, duties on such items were reduced to 27 percent. Customs bands were streamlined to four: 0, 10, 15 and 27 percent. In the 2002 budget, however, tariffs were increased from 10 to 27 percent for beef and chicken meat and knitted or crocheted fabrics. Currently the standard rate of duty is 10 percent, although most production inputs are subject to a rate of 3 percent. Intermediate goods are generally subject to a 15 percent import duty. 2.19 The breakdown and trends in revenue are shown in Figure 2.4 and Table Al. As the general level of tariffs was reduced in Fiji, import duties declined from 30 percent of 42 total revenue (including grants) in 1989 to 19 percent in 2001. Revenue from the VAT grew quickly to 24 percent of total revenue in 1993: the first full year after the introduction of the VAT and stood at 28 percent in 2001. In the same period, total government revenue (including grants) has averaged around 26 percent of GDP. Other notable developments have been a decline in personal income tax from 23 percent of total revenue in 1989 to 15 percent in 2001, a small decline in the share of excise duties, the replacement of the turnover tax and a decline in grants. Figure 2.4: Fiji - Trends in Revenue 60 - > b~~~~~Icome, Tax 30- 0~ ~~~~~~~~~~~~~~~~0 U /VA 0 H 1989 1991 1993 1995 1997 1999 2001 2.20 During the mid to late-1990s major tariff and tax refor-ms were undertaken in Samoa. The VAGST introduced in 1994 laid the foundations for significant reductions in tariff rates. Import duty reductions were made in 1990, 1994 and 1996 but the main tariff reforms were associated with the 1998-1999 and 1999-00 budgets. The reforms in the 1998-1999 and 1999-00 budgets included reducing tariffs on most business inputs from 20-42 percent to 10 percent and on most consumption goods from 50-60 percent to 10-20 percent. Tariffs on basic foods, medicine, fertilizer and similar items were unchanged in the range of 0-10 percent. 2.21 A 10 percent VAGST is payable on all consumption but with exemptions for local unprocessed primary production, financial services, bus and taxi services and since 1998- 99 for businesses with a turover of less than S$52,000. A zero rating applies for exports and hospital and educational services. 43 Figure 2.5: Samoa - Trends in Revenue 60 - a ' 30- Cd Trade Taxes VAGST 1989 1991/92 1994195 1997/98 2000/01(est) 2.22 Reflecting the reforms implemented in the -last 10 years, trade taxes have declined from over 30 percent of total revenue in the early 1990s to an average of around 12 percent of total revenue in recent years (see Figure 2.5 and Table A.3). Revenue from excise taxes increased from an average of around 9 percent during the 1989-1997 period to 19 percent of total revenue in 1999-00 and 2000-01. This change was associated with a sharp rise in excises on imports in 1998--99 and a further increase in 1999-00. This corresponds approximately to the reduction in import duties that occurred in those years. Increase in domestic excise revenue was much smaller. Over the same period there has been a reduction in the contribution from income taxes, currently around 14 percent of total revenue. 2.23 Vanuatu introduced a VAT in mid-1998 combined with a reduction in import duty rates, the removal of a custom services charge of 5 percent on imports, and the rationalization of a number of other minor revenue measures (e.g. turnover and hotel taxes). The tax and duty reform entailed a reduction in total import duties of 25 percent over two years. Import duties were reduced to a band within zero to 30 percent and simplified to a total of eight categories. By 2001. the VAT was collecting as much revenu,e as import duties.and together they accounted for around 70 percent of revenue. However, total revenue from VAT and import duties combined as compared with the revenue measures they replaced, fell from 20 percent of GDP in 1997 to 18.5 percent of GDP in 2000. The ratio is projected to remain below 19 percent of GDP over 2002 and 2003. The revenue shortfalls have led to budgetary problems which were particularly acute in 2001 and required the introduction of tight expenditure constraints. 44 Figure 2.6: Vanuatu - Trends in Revenue 60 - Trade Taxes e 30- So ~~~~~~~~~~~/ o / VAT TumoverTaK . T 0 1989 1991 1993 1995 1997 l999p 2001p 2.24 The reasons for the revenue shortfall are not entirely clear. Contributing factors are believed to have been a slow down in economic activity, an underestimation of the tax rates required for a revenue neutral reform, some compliance problems with the VAT (at least initially) and import duties, and the loss of import duty revenue as resources were re-directed to the VAT from the Customs Service. The commencement in 1998 of the bilateral trade agreement with Fiji also contributed to the revenue shortfalls due to foregone import duties. 4.2 Issues in VAT implementation Zero-rating and Exemptions 2.25 The major problem with the implementation of the VAT in Fiji related to the administrative resources required to deal with refunds for zero-rated sectors. Agriculture, sugar producers and small-scale enterprises were zero rated, meaning they did not levy VAT on their sales but could claim a refund for VAT paid on inputs.3 The tax system had to accommodate thousands of claims (for VAT refunds) which congested the administration system and entailed significant delays. There were also substantial non- compliance problems. In the first year of the VAT it was estimated that only 30 percent of taxpayers were complying with the law and at the end of 1993 VAT arrears had accumulated to about one-third of total collections for the year. In general, compliance has been better at the ports because VAT is much easier to collect on imports. 3The standard terminology is that zero rating means that sales are taxed at 0% but that credit is given for any sales tax paid on inputs. In contrast exemption means that sales are not taxed but no credit is given for any sales tax paid on inputs. The latter is obviously easier to administer but entails taxes on business inputs. 45 2.26 The exemption from VAT of basic food items is practiced in many countries in order to ameliorate the distributional consequences of the tax. It is instructive to review Fiji's experience in this regard. In the 2000 budget, some basic food items-rice, flour, tea, edible oils, tinned fish and powdered milk-were exempted from VAT. However, effective January 2001 the VAT was reinstated. In the government's view the objective of helping the poor is more effectively achieved through altemative, less costly, measures. According to the government, the policy of VAT exemptions on basic food items had several shortcomings:- * The policy is not well targeted because all consumers who buy the items benefit, not just the poor. In its view a better alternative would be to assist the poor directly through special income support and poverty alleviation projects. * The estimated revenue loss to the government was of the order of F$ 40 million or nearly 20 percent of VAT revenue in 2000 * Exemptions increased the cost to businesses of maintaining dual systems to cover VAT and non-VAT items; and * The exemptions complicated the ability of the Inland Revenue department to collect VAT, increased the opportunity for tax evasion and shrank the tax base. At the same time as the exemptions were removed, import duties on these basic foods were reduced in order to cushion the impact on prices. 2.27 In Samoa, the implementation of the VAGST greatly stretched the capacity of the Internal Revenue Department. As in Fiji a major problem arose in dealing with refunds of VAGST for sectors that have zero-rating. However, the problem does not appear to have been on the same scale as in Fiji where relatively more entities were zero-rated. Businesses have advised that government officials have not always appreciated the administrative and financial costs that they have to endure when refunds are slow to come in. The VAT Threshold 2.28 Another issue related to the VAGST was that of a threshold firm size above which registration for the tax is compulsory. If it were not for the costs of administering the VAGST (incurred by the government) and of complying with it (incurred by the businesses), the best threshold would be zero (Ebrill et al 2001). The need for a threshold arises from the advisability to forsake some revenue in order to save on collection costs. In Samoa, to reduce the costs of administering the VAGST system for both the authorities and small businesses, an exemption for small businesses with a turnover of less than S$52,000 (about US$15,000 at the average exchange rate for 2001) was introduced in 1998-99. The choice of threshold was supported by the fact that the largest 10 percent. of the registrants were responsible for 80 percent of revenue. According to the authorities in Samoa, deploying scarce administrative resources more effectively on larger taxpayers is a strategy that has worked well. The threshold in Fiji is F$30,000 (about US$13,000) for supply of goods and F$15,000 for supply of services. 46 VAT on Imports 2.29 As in many other countries a large proportion (70 percent in 2001) of Vanuatu's VAT revenue is collected at the port of entry by the Customs Service.4 The remaining 30 percent is collected by a separate VAT unit. In 2001, the top 100 taxpayers yielded 92 percent of the VAT and import duty collections at the port of entry compared with 49 percent of revenue collected by the VAT unit. Thus targeting larger tax payers and strengthening enforcement at the port of entry are important priorities for increasing revenue. Similar arguments apply for both Fiji and Samoa. VAT Audits and Collection Costs 2.30 In Vanuatu, the costs of collecting VAT were around 6 percent of VAT revenue whereas the costs of collecting import duty were around 1.4 percent of total duty. As in other countries, one of the reasons for the relatively high ongoing cost of the VAT is the need to reinforce compliance through an extensive audit program. Vanuatu is auditing all registered VAT payers on a 36-month cycle. This means that each of the 1,100 or so registered VAT payers can expect at least one audit very 24 months, where the nature of the audit could range from a simple verification of key figures to an intensive review of company accounts. Much of the audit work was initially contracted out to local accounting firms.5 Tougher penalties have been implemented to enhance compliance. 5. Reform Considerations in the Context of Trade Liberalization 2.31 PICTA is unlikely to have an adverse impact on tariff revenues for Fiji. Fiji is considered to have greater capacity to compete with other PICs and to gain by increasing its exports to other PICs and experience little diversion of its imports to other PICs. 2.32 In fact, Fiji is probably one of the PICs that can benefit by moving at a faster pace to integrate itself more closely with the global economy. In this context, there would be a need to explore alternative revenue sources given that import duties still comprise about 17 percent of total revenue. As the contributions from VAT and income tax are already relatively high (see Table 2.1) and the contribution from excise taxes is relatively low there is likely to be scope to raise more revenue from excise taxes. The excise taxes in Fiji on commodities such as alcohol and tobacco do not apply equally to domestic production and imports. hnports are charged a duty which is higher than the excise paid on domestic production. In the 2002 budget, both these charges were raised and the differential was maintained. Removing the differential would offer one way of increasing revenue if trade liberalization beyond PICTA is considered. 4 This refers to the gross collection and not the net payment after refunds. Importers are able to claim the VAT paid on imports as a credit against VAT paid by the domestic buyer of the imported items. The gross collection is equal to what would have been collected if the tax was payable at the port only. 5 The presence of such firms, which were largely established to meet the needs of the Offshore Finance Centre, provided an important resource to be called on in implementing the VAT. Conflict of interest controls were put in place to prevent a private accounting firm undertaking a VAT audit of existing or recent clients, and the involvement of these firms is now being scaled back as audit skills are developed within the Government's VAT Unit. 47 2.33 Income earned from exports is currently tax free, however, this treatment is to be phased out so that all export income will be fully taxed by 2009.6 This will also allow Fiji to fulfill its commitments under the WTO, as the exemptions are not in conformity with WTO obligations. The announced policies to discontinue the Tax Free Factory and Tax Free Zone schemes should also assist in raising income tax revenues. Other options include more effective enforcement of the taxation of fringe benefits. 2.34 There are numerous public enterprises in Fiji that do not pay income tax. This is considered to be an important distortion, particularly where there is scope for direct competition with the private sector. However the 2002 budget indicated that changes will be made to the Public Enterprises Reform Act to address this problem. 2.35 The government estimates that by further strengthening the Fiji Revenue and Customs Authority (FRCA), particularly the VAT unit, and simplifying the customs legislation, compliance could be significantly improved. Strengthening .-institutional capacity would involve upgrading of the tax information system, training, and reviewing its remuneration structure. The government plans to collect F$50 million out of an expected stock of arrears of around F$130 million (or 13 percent of budgeted revenue) in 2002. In strengthening compliance, the government expects to collect an additional F$40 million. 2.36 In the case of Samoa, Fiji is the only PIC supplying significant imports, with a share of 9.4 percent of total imports in 2000. Other PICs supply less than 0.2 percent of total imports. The°se figures suggest that PICTA is likely to have a very small impact on tariff revenue for Samoa. For instance, if Fiji could double its exports to Samoa the maximum impact would only be about 2 percent of total revenue. Even if Samoa was to completely eliminate its tariffs the revenue-raising task is considered to be manageable given the relatively low contribution from income tax. The highest marginal tax rate is 29 percent, there is limited enforcement of the taxation of fringe benefits, and income taxes only constitute 14 percent of total revenue (about 4 percent of GDP). 2.37 There is scope to address a decline in tariff revenue by removing the tax exempt status of primary production. The exemption from taxes of agriculture and fishing may have been valid on equity and administrative grounds in the past, but the recent changes to raise the tax free threshold to S$10,000 which is about twice the per capita income make such an argument redundant now. The exemption for primary production means that high income commercial farming and the booming export fishing industry do not pay any income (or royalty) tax. These exemptions are difficult to defend both on efficiency and equity grounds. 2.38 There is also the potential for raising revenue through increasing the taxes on luxury products. The 1998-99 budget entailed substantial cuts in excise taxes and import duties on luxury products. For example the combined duty and excise on a new car with a landed value of S$20,000 was reduced from S$25,000 to only S$4,000 and the 6From 2003 to 2004, 25 percent of export income will be subject to tax. From 2005 to 2006, 50 percent of income will be taxed. From 2007 to 2008, 75 percent of income will be taxed, and in 2009 all export income will be fully taxed. 48 combined duty and excise on a television set worth S$500 was reduced from S$370 to only S$8 (ADB 2000). 2.39 Vanuatu's experience with the MSG Trade Agreement and the Bilateral Trade Agreement with Fiji implemented in 1998 provide an early indication of the potential impact of PICTA. The Government has estimated that the revenue loss in the early years of the full operation of the MSG Trade Agreement at US$0.5 million. This is expected to rise to about US$1.5 million by 2005. Much of the revenue loss to date has arisen from imports of tinned meat, tinned fish and cement from Papua New Guinea and Fiji. Imports from MSG partners now account for almost 70 percent of tinned fish and tinned meat imports and approximately 50 percent of cement imports. The overall revenue loss estimated for 2001 is small at 4 percent of total tariff revenue or US$0.6 million. 2.40 Adoption of PICTA would see the removal of import duties on a broader range of goods than covered under the existing trade agreements. The commodities most likely to be.a potential source of revenue loss are beer7 (which accounts for 4 percent of tariff revenue), some tobacco products (tobacco products provide approximately 15 percent of tariff revenue) and petroleum products (30 percent of tariff revenue). These products are normally subject to excises, and it would be appropriate for Vanuatu to convert the existing tariffs into excises at a similar level. This would offset the potential loss of more than 30 percent of current tariff revenue. 2.41 In late 2001, plans were developed by the government to expand the excise regime with a view to levying duties on 'sin' goods. Implementation was delayed due to difficulties in putting new legislation in place. The potential loss of revenue under PICTA and PACER will add to the need to implement this expanded excise regime, and the introduction of such reforms would appear to be the minimal response to the adoption of PICTA and PACER. It would also be reasonable to extend the excise regime to cover luxury goods at a high rate as a means of increasing the progressivity of the tax system. 2.42 For non-excisable goods, the combination of relatively low tariffs in Vanuatu and the existing tariff preferences provided to MSG partners means that PICTA may have little impact on revenue. 2.43 While the tax reforms of the Comprehensive Reform Program saw a reduction in taxation of business, import duties remain on some inputs to business and this reduces the international competitiveness of the economy. For example, there is: a 5 percent tariff on fertilizers and hand tools; a 15 percent tariff on most steel products including barbed wire; a 25 percent tariff on measuring instruments (e.g. weighing machines) typewriters and cash registers; and a 35 percent tariff on photocopiers. While a duty drawback scheme is in place, it does not extend to all business inputs or all businesses. 2.44 Even if the revenue loss for Vanuatu was small, there is value in continuing tax reforms in order to increase the productive potential of the economy. Immediate priorities include the extension of the excise regime, the removal of import duties on 7Alcohol and tobacco are currently excluded from PICTA. These revenue losses would occur only if they are included, as envisaged in the future. 49 business inputs, and a review of the relatively high business license fees on small to medium businesses. It is also important that proposals to establish additional business taxes do not impose inappropriate costs on business.8 2.45 The widespread liberalization of trade envisaged under PACER can be expected to see the loss of much of the tariff revenue collected on non-excisable goods. This is in the order of 45 percent of current import duty collections. Such liberalization would probably require a considerable broadening in the tax base and a higher VAT rate. 2.46 The downward trend in Vanuatu's revenue is against a background of a relatively low level of internal revenue. The Government established a Revenue Strategy Committee in 1999 with a view to exploring means of lifting the ratio of total government revenue (excluding grants) to GDP from the existing level of 22 to 23 percent up to 27 percent. Publicly canvassed options for tax reform include: an extension of the current excise regime (that currently only applies to domestically produced beer); the introduction of a business tax to cover the retail sector and local companies; the introduction of a payroll tax; the adoption of a presumptive tax on smaller operators (there is currently a threshold exemption level for VAT for smaller operators); and the review of existing minor taxes and charges (e.g. license fees, road taxes). Additional options that could be considered include an increase in the VAT rate above the current level of 12.5 percent and the adoption of income taxes. 2.47 Vanuatu does not levy an income tax and is reluctant to do so given the potentially adverse impact this may have on foreign investment and its status as an offshore business center. There can be valid economic arguments for not levying direct taxes on income.9 It may be more efficient to raise tax revenue from taxes on consumption. However, for Vanuatu the absence of income taxes makes it more difficult to achieve a reasonable progressivity in the tax system. If the current system is seen as inequitable, the problem could potentially be rectified by taxing luxury goods at a relatively high rate or introducing taxes that target higher income groups. For example, a land tax on higher valued land could be an efficient way to increase tax revenue from higher income groups. 6. Kiribati and Tonga: Potential Impact of PICTA and Policy Response 2.48 This section examines the potential impact of PICTA on two of the more exposed countries. The issues analyzed in both cases are, the potential impact of PICTA on tariff 8 For example, the proposed payroll tax would be paid by business, but it can be expected that it will be passed on to employees. If so, it would have little, if any, impact on Vanuatu's international competitiveness. A small turnover tax on the finance sector may also be an efficient way to tax the profits of the~ sector. But any tax that adds to the costs of a business that has limited capacity to pass the tax on or a poor capacity to pay should be avoided. For example, additional licence fees for small manufactures would be undesirable. 9 For example, most income taxes lead to the double taxation of savings and this distorts savings decisions. And in the absence of an imputation system, company tax systems lead to the double taxation of dividends and distort the incentive to invest and engage in business. 50 revenues, the key issues raised by the current tax and tariff systems, and the reform considerations in the face of trade liberalization. 6.1 The Potential impact of PICTA 2.49 The approach adopted to estimate the potential revenue loss identifies goods that have relatively high tariffs in Kiribati and Tonga, that are produced in other PICs to a substantial extent, and that are considered likely to be candidates for trade diversion. 2.50 Kiribati sources nearly 20 percent of its imports from PICs (almost all from Fiji) and some 16 percent of tariff revenue was collected on these imports. Of the imports recorded from the PICs, some are re-exports from other countries. For example, almost $6 million of imports recorded from the PICs is fuel from Fiji. This fuel is simply shipped via Fiji, and would not qualify for a tariff concession under PICTA as it originates in a non-PIC (mainly Australia and Singapore). Excluding fuel, PICs account for 10 percent of Kiribati's imports and 10 percent of duty collections. This means that the immediate loss of revenue from PICTA would be small, probably less than $1.5 million. 2.51 But there is potential for imports from the PICs to replace other imports and for the share of PIC trade to grow. This is most likely for processed food (e.g. tinned meat and fish), beverages, paper products, toiletries, garments, cement, and timber products as these goods are already produced in the PICs. 2.52 It is difficult to predict the long term potential to switch to PIC products. Wholesalers contacted in Kiribati advised that they would certainly look to import more PIC goods if they were much cheaper, but there were mixed views on the likely extent of substitution. Factors highlighted were the limited product range in the PICs, that quality tended to be noticeably lower, and that consumers would initially at least be reluctant to switch from known food brands (i.e., they like the taste of their current product). However, competition amongst wholesalers and the low incomes in Kiribati can be expected to see substantially lower priced PIC goods gain market share and for PIC producers to take advantage of their favored market position. 2.53 An issue for Kiribati is the imminent opening of a petroleum refinery in PNG, the Napa-Napa oil refinery. This is intended to supply both domestic production and exports. Petroleum products from PNG would warrant a tariff concession and have the potential to replace all fuel currently imported from Australia and Singapore. Fuel alone accounts for 6 percent of Kiribati's tariff revenue. 2.54 Most of Kiribati's beer is imported from Australia, with only a small share from Fiji, and accounts for 12 percent of tariff revenue. But all of the larger PICs produce beer and it is reasonable to expect a major shift to PIC-beer if provided a substantial price advantage (the tariff on beer accounts for around 25 percent of the retail price). Since alcohol has been excluded from the current version of PICTA these revenues will only be impacted if the agreement is extended to cover alcohol. 51 2.55 Sugar, beef, tea, coffee, chicken, eggs, and soft drinks are other goods produced in the PICs that are subject to high tariffs in Kiribati and are candidates for trade diversion. Imports of these goods from non-PICs accounted for 16 percent of tariff revenue in 1999. While Kiribati has only a small priyate sector, there are a number of activities that could be in strong competition with other member countries under PICTA. There are local producers of eggs, chicken, garments, boats, biscuits, nails, ice cream, canvas products, aluminum products and flip-flops (thongs), and a local producer of copra products is scheduled to commence soon. Most local producers are protected by substantial tariffs and- could find it difficult to compete without this protection. Under PICTA they may cease operation or contract, leading to a loss of revenue for the Kiribati government. 2.56 The garment industry illustrates this point. It employs around 100 people protected by a 60 percent tariff. Made-to-wear garments, that account for around half of total sales, are not readily exposed to competition from imports. But the other main sales, school and company uniforms, could be readily produced in Fiji. The larger, lower cost producers in the Fijian industry may displace local producers, reducing tariff revenue on imported inputs, and company and personal income tax collections. 2.57 The foregoing discussion suggests that consistent with the earlier studies of revenue loss it is quite possible for Kiribati to forego 20 percent of its tariff revenue, but this could be made up through other means as elaborated below. 2.58 In Tonga, total imports originating in PICs were 11 percent of total imports in 2000 and accounted for 5.3 percent of total duty collections. As with Kiribati oil imports routed through Fiji need to be excluded. Excluding these, total imports from PICs in Tonga are about 8 percent of total imports and 2 percent of total duty. Fiji is the dominant PIC supplier with Samoa supplying a.share of only 0.3 percent of total imports. This suggests that the initial revenue loss from PICTA would be small. However, it is likely that trade diversion would develop and increase the revenue loss considerably. 2.59 Fiji is a significant supplier in virtually all the major import categories. It supplies about 16 percent of food and beverage imports, 10 percent of wood and wood products; 19 percent of textiles and footwear; 12 percent of basic building materials; 6 percent of machinery other than transport equipment; and 15 percent of other, miscellaneous manufactured goods. 2.60 The category animal and vegetable products and food and beverages, constitutes nearly 33 percent of total imports, but 64 percent of duty collections. This large share reflects the importance of cigarettes as a revenue source accounting for about 24 percent of total duty. Beer and other alcoholic beverages provide 12 percent of total duty. 2.61 A precise estimate of the potential revenue loss from trade diversion would require a case-by-case assessment. However, a review of the detailed data suggests that the loss could easily range up to 15 percent of total tariff revenue. 52 6.2 Current Tariff and Tax Systems in Kiribati and Tonga 2.62 Import duties are the main source of tax revenue in Kiribati, followed by company tax, personal income tax, and a hotel duty. There are no other taxes. Import duties comprise a tariff and an import levy. 2.63 The current tariff structure is based on 16 different ad valorem tariff rates for 1,953 items, ranging from zero to 100 percent. In addition to the ad valorem duties, there are 72 items that are subject to specific duties. When converted to ad valorem rates, specific duties range from 1 percent to 305 percent. In addition to the customs tariffs, there is an import levy which is applied on a specific basis on almost all products. It is estimated to be equivalent to 2.8 percent on average in ad valorem terms. The levy applies to all goods other than four petroleum products. The average unweighted tariff was calculated at 22.4 percent and the import-weighted average tariff is 24.9 percent inclusive of the import levy. 2.64 The tariff system collects around 50 percent more revenue than the tax system. It was the subject of a major review in 1997 and the findings remain relevant (World Bank 1997). The main concerns are a large number of tariff rates, the use of FOB rather than CIF in valuing imports, exemptions provided to certain government enterprises, the widespread use of specific tariff rates, the protection provided for certain local producers, and the absence of an established system of duty drawback. There are also concerns that the tariff system suffers from similar avoidance and evasion problems as the tax system with the potential for incorrect valuations of imports being the main concern. 2.65 Kiribati uses progressive company tax rates which may distort business structures. This is because they may encourage businesses to split into multiple companies to take advantage of the lower tax rates applicable to lower levels of profit. Other areas of concern are the double taxation of company income for foreigners, the discretionary nature of certain company tax concessions (provided under the Pioneer status), limited provisions for deducting interest and the short duration of the loss carry forward provision. 2.66 One of the features of the Kiribati revenue system is the absence of an excise tax. These are taxes applied equally to imports and domestic production, and are used in many countries to raise revenue, deter the consumption of 'sin' goods and achieve equity objectives by applying them to luxury goods. Goods normally subject to excise are tobacco products, alcohol, soft drinks, motor fuel and (in some cases) motor vehicles. Luxury goods are often subject to high excise rates. 2.67 The replacement of the current tariff on these goods by an excise would do much to preserve Kiribati's revenue and minimize the trade diversion-induced welfare loss that Kiribati may incur under PICTA. Excisable goods account for slightly more than 40 percent of Kiribati's import duty collections. There is also a case to consider imposing an excise at a higher rate than current tariffs on some goods e.g., the average tariff on fuel is less than 20 percent and fuel excises are normally considerably higher. A well designed excise (with rates set to both raise revenue and achieve equity objectives) could 53 potentially preserve about 50 percent of Kiribati's current tariff revenue, and this revenue would be preserved even if more general trade liberalization was pursued. 2.68 It is important to recognize that Kiribati's largest single revenue item is the collection of license fees from foreign-operated fishing vessels. The profitability of such operations in the region and the appropriate level of license fees have received considerable attention But the scope to tax more of the economic rent (surplus) associated with fishing in the region needs further study. It is important that Kiribati maximizes its revenue from its marine resource while ensuring sustainable use of the resource.10 2.69 Tonga too is heavily reliant on trade taxes and the tax system is characterized by economic distortions. Most tariffs are 15-25 percent with an additional ports and services tax (effectively also a tariff) of 20 percent. The duty on vehicles is 45 percent and on vehicle parts 35 percent with an additional ports and services tax of 20'percent. The Tongan tax and tariff system has been the subject of a number of authoritative reviews (IMF 1990, ADB 1996) and the government has been aware of numerous problems with the current system for some time. 2.70 The basic features of the current tax system are as follows. There is a corporate income tax with differentiation between foreign and local companies, preferential treatment for designated export-oriented industries and progressive rates for most entities. Local entities pay up to 30 percent and non-resident companies pay up to 42.5 percent. There is a low personal income tax rate of 10 percent, with extensive deductions including allowances for housing and loan repayments, making the effective tax rate even lower. Dividends, interest and royalties are also subject to a withholding tax of 15 percent and income from leasing land is subject to a withholding tax of 3 percent. 2.71 A retail sales tax of 5 percent applies for imported and locally-produced goods. There are exemptions for goods for personal and government use, and for building materials, wheat and flour. Smaller entities only pay sales tax on their inputs as there is no system for collecting tax on their sales. Entities that meet a T$lmillion turnover threshold are'registered. They are only required to remit sales tax to government after the tax has been levied at the point of sale. Other entities are required to pay when goods are cleared through customs or purchased domestically. Sales from one registrant to another are exempt from the sales tax. There is also an accommodation and restaurant services tax of 2.5 percent. 2.72 In 1999 the Ministry of Finance created a Tax Policy Committee to address the problems with the current tax system and propose options for the Government to consider. A tax reform paper has been prepared and presented to government for its consideration. Major changes being contemplated include reducing the reliance on trade taxation, simplifying the tariff structure, simplifying and aligning individual and corporate income tax, increasing the degree of progression in the income tax, 10 Some issues related to fishing license fees were covered in the World Bank's Pacific Islands Regional Economic Report (2000), Cities, Seas, and Storms: Managing Change in Pacific Island Economies. 54 implementation of fringe benefits taxation, abolishing discretionary exemptions and expanding the sales and excise taxes. Under the new arrangements, exports would be zero rated and exporters would be able to claim credit for all duties, sales tax and excises paid on business inputs. 2.73 In the past there has been substantial scope for evasion and manipulation of the incentives schemes that exist under the Industrial Development Incentives Act. It is now recognized that the Act has not been effective in promoting industrial development, has entailed significant loss of revenue, and has been abused by some individuals by not operating businesses in accordance with their original applications. In addition, the scope to use discretion in providing exemptions has been seen as unfair and open to abuse. The government is in the process of taking legal action in relation to a number of alleged instances of fraud in relation to the Act. The Act was suspended in October 2001, and is to be repealed as part of the tax reform. 6.3 Administration and Capacity Constraints 2.74 Perhaps the most pressing issue facing the tax system in Kiribati is one of administration. The Taxation Office employs only two qualified accountants. This skill constraint, combined with the small number of staff, is hampering the ability of the Taxation Office to administer the company and personal tax systems. The Office is responsible for 16,500 individual taxpayers and 102 company taxpayers, but checks of tax returns are very limited. For example, in recent years only 2 company audits and 7 individual audits have been conducted (covering $150,000 in tax payments, or around 1.5 percent of annual tax collections). Under these conditions there is potential for considerable tax evasion and avoidance, particularly in the case of companies and non- government employees. The Taxation Office has already identified US$2.2 million in arrears (equivalent to around half of the annual collection), 21 delinquent company taxpayers and 495 delinquent individual taxpayers. 2.75 In Tonga there are weaknesses in administrative and enforcement capacity for all major taxes. Legislative provisions are needed to facilitate compliance by taxpayers and protect their rights. These are to be addressed as part of the tax reform package. It is estimated that revenue amounting to between 3-5 percent of GDP is lost through weak administration. 2.76 There are substantial capacity constraints in the income tax and sales tax administrations. Currently there are around 60 staff in inland revenue, but only one graduate and two staff with diploma qualifications. Computers are not linked and many key staff do not have adequate access to computing facilities. Staff are located in different locations and the administration generally finds it difficult to fulfill its current functions. There is a concern that there are insufficient audits and existing legislation needs to be revised to improve compliance. The tax reform package will need a major effort in strengthening institutional capacity if it is to be successful. It may not be feasible to implement a comprehensive package without a staged and selective approach to its key elements. 55 6.4 Considerations for the Introduction of a Broad Based Consumption Tax 2.77 The administrative and capacity constraints noted above would hamper the management of any tax system, but are likely to be an even greater hurdle for the introduction of a VAT or VAT-like tax. In addition, there are other issues that need to be considered before the introduction of such a tax. 2.78 Kiribati has a very narrow economic base. Despite the above concems it is unlikely that there are substantial distortions created by the tariff system, as there is little domestic production. Under these conditions, a tariff is akin to a consumption tax, and one which is collected very effectively. It is reasonable to argue that given the high import dependence of the economy and institutional constraints, consideration should be given to converting the tariff to a simple sales tax levied at ports and at the final point of sale for selected services such as utilities and accommodation. This should be manageable from an administrative perspective. If the administration weaknesses could be more effectively addressed, an altemative would be a sales tax at the ports and on business registrants above a certain size at the point of sale, with registrants being exempt from paying taxes on inputs or able to claim a credit when they remit taxes on their sales to the authorities. 2.79 Based on the latest import and duty statistics, the same level of revenue as currently collected from import duties on non-excisable goods could be collected by a uniform sales tax of the type described above. The rate of this sales tax would be of the order of 12 to 15 percent, depending on whether it is levied at the wholesale or retail stage. This estimate excludes the main imports normally used in production, which ideally should be free of taxes and tariffs. Alternatively, a uniform 10 percent sales tax could be implemented with a low, uniform tariff of around 5 percent for revenue raising purposes. 2.80 Such a revenue system would benefit Kiribati by removing some of the problems of the current import duty system, maintain existing current revenue levels and avoid altogether or minimize the potential for trade diversion under PICTA. The absence of a tariff or a low tariff would also negate the need to develop potentially complex rules of origin required under preferential trade agreements. The key issue for Kiribati to consider would be the appropriate form of a sales tax. Simple tax systems are attractive, and a full VAT or GST is probably too complicated at this stage. 2.81 Implementing and administering a new- sales tax would require substantial efforts at improving capacity. This would include, policy development, legislative amendments, education of taxpayers, upgrading of the computer system (computerization is incomplete at this stage), the training of staff and revised administrative procedures. Such additional responsibilities can be expected to divert resources from the oversight of the company and personal tax systems, and the upgrading of accounting skills. This could weaken the overall tax system and collections. It may instead be desirable to levy a single stage sales tax at the port, at the factory gate for manufacturers and at the retail level for certain services. The exact form of any future sales tax requires careful thought and investigation taking into account the institutional constraints. 56 2.82 In Tonga, there has been considerable consultation about the'tax reform package, but it has not yet been finalized. Further awareness campaigns are being developed for the general community. It appears that the main aspect of concern to both government and business relates to the expanded sales tax. It is proposed that this will entail a broader base (covering all goods including those for personal use, and services above a certain threshold) and a higher rate. The exact form of the tax has not yet been announced but is likely to have VAT-type features. Entities above a certain threshold would be able to claim credits on taxes paid on inputs, and exporters would be zero rated i.e., not pay taxes on inputs or outputs. The key trade-off is the potential for such a tax credit mechanism to enhance economic efficiency versus the additional administrative costs for both government and business. 2.83 From the government's perspective the main concern appears to be the community attitude to higher sales tax on a wider range of products, including those for personal use, and for services, including utilities. From the business perspective the main issue relates to the credibility of the input tax-credit concept being effectively implemented, and its compliance costs. Various business, entities that were recently consulted did not consider the scope for claiming tax credits as a significant benefit. This was because of the very limited extent of vertical economic relationships, and associated value added. Exceptions to this would be in the export sector where there is no scope to pass on taxes through higher prices. 2.84 The contribution of the manufacturing sector to GDP in 2001 was only 5.2 percent -- virtually the same as in 1975. The manufacturing activities are generally basic food or raw material processing and simple assembly. For example, bakeries are an important part of the manufacturing sector. There is a brewery, several firms making concrete blocks or other basic building materials, several firms making furniture and one or two firms involved in activities such as paint mixing. The main commodity exports are squash and fish. Given the structure of the economy it does seem reasonable that the main issue with respect to tax credits would relate to the export sector. 2.85 Administrative costs of collecting the sales tax are currently very low. In 1998/99 administration expenditure in relation to sales tax was 1.7 percent of revenue compared with 6 percent for income tax and 1.2 percent for customs.revenue. For sales tax, the relatively low expense-to-revenue ratio can be explained by the very small number of entities, currently twenty, that meet the threshold for paying the tax after the goods are sold. Rest of the tax is collected at the port of entry with the assistance of the Customs Department. Under the proposed tax, thresholds for the tax-credit mechanism should help to keep collection costs at acceptable levels. 57 7. Key Findings and Lessons for PICs 7.1 PICTA and PACER Create a Needfor Tax Reform 2.86 Although the potential extent of trade diversion effects is unclear, it seems likely that these would occur in the case of Kiribati and to a lesser extent Tonga. These countries are heavily reliant on trade taxes as revenue sources and do not have a VAT- type tax in place. Consideration of a broad-based consumption tax would go hand-in- hand with the adoption of PICTA. 2.87 PICs that already have a VAT may initially be able to compensate for any revenue losses through small increases in the tax rate, through the use of excises or through enhanced efforts to improve compliance in the tax system. The Solomon Islands and Tonga may be able to use an increase in the rate of their simpler sales tax as a short term means of offsetting any revenue loss under PICTA. 2.88 However, wider-ranging reforms would be required to exploit the benefits of trade liberalization beyond PICTA. The desirable features of a revenue system for the PICs would include: * A simple and broadly-based progressive income tax that avoids the double-taxation of dividends and provides early and full deductibility of costs. If such a tax is not in place, alternatives may be required to ensure appropriate progressivity in the tax system. Fringe benefits should be taxed and exemptions from income tax should be avoided. * Reasonably high excises on 'sin' and luxury goods applied equally to domestic production and imports. This would be an effective way to recover the revenues that would be lost through reductions in tariffs on these goods. Taxes on luxury goods could also be a way of introducing progressivity in the tax system. * For countries that have not already done so, a broadly-based sales tax applied at a uniform rate of the order of 10-15 percent. Options include a VAT or a simple sales tax levied at the ports and at the point of sale for selected services. Exemptions are likely to be warranted for small businesses below a threshold level to help keep administration costs for government and business at acceptable levels. Exemption means that businesses do not charge taxes on their outputs or remit tax to the authorities, but would pay taxes on their inputs. * Either no tariffs or a low and generally uniform tariff (applied on CIF values) on consumer goods for, revenue-raising purposes. Greater uniformity in tariff rates simplifies administration and minimizes the potential for evasion through incorrect valuations. * For some countries in the region, the taxation regime for resource-based activities, e.g. fishing, needs to be improved so as to more effectively capture the rent from natural resource exploitation. 2.89 Most PICs have steadily moved towards such a desirable revenue system or signaled their intention to consider reform along the lines outlined above. PICTA and 58 PACER reinforce the importance of adopting such a revenue system, as well as tend to highlight the cost of high tariff regimes, and tax systems that encourage inefficient activities. 2.90 Tariffs are used widely in the region to protect local industry (e.g. producers of garments, eggs, chicken and biscuits). Some countries also use tax systems to encourage local production, for example by providing tax exemptions. Adoption of the desirable tax and tariff systems described above would see the removal of most if not all of this protection and assistance. A strong case can be made against a general policy of providing protection. But if regional governments do wish to protect local industry and reform the tax and tariff systems, an alternative means of providing protection would be needed. It would be preferable to provide transparent subsidies to assist industries rather than using the tax and tariff system as a means of providing support. 7.2 Tailor the Tariff and Tax System to Need and Local Institutional Capacity 2.91 The difficulties faced by most countries in the region when implementing a VAT and the capacity constraints that limit the sophistication of the revenue system highlight the importance of tailoring the revenue system to local institutional capacity. 2.92 This issue is most critical in the consideration of a desirable sales tax. Three of the PICs already operate a VAT, and there may be a tendency for the other PICs to pursue similar reform. There is a trade-off between efficiency, equity and administrative simplicity in tax systems. The PICs yet to put a VAT in place are characterized by a relatively narrower economic base and greater institutional constraints. In these countries the potential gains from the relatively sophisticated VAT are likely to be small and taxes that are simpler to administer may be preferable. An option that should be carefully considered is the adoption of a simple sales tax initially with a view to graduating to a VAT when institutional capacity is adequate. Alternatively, a VAT could be applied with a relatively high threshold for exemptions. Firms above the threshold would be formally registered and be required to collect tax on their outputs and rernit it to the tax authorities after deducting taxes paid on their inputs. All other businesses would pay tax when they purchased their inputs, but not be required to collect tax on their outputs and not be required to submit VAT returns to the authorities. The VAT could be levied at the port of entry and at the point of sale of business registrants above a certain threshold. 59 Table A.1 Fiji Government Revenue 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 (percent of total revenue and grants)* Tax revenue 80 80 81 83 83 83 84 80 79 82 Income tax 32 29 29 31 31 28 28 26 30 28 Personal 18 14 16 16 17 16 17 15 15 15 Company 14 14 13 15 14 12 11 11 15 14 Tax on goods & services 23 30 31 30 31 32 33 33 29 35 VAT 13 24 24 24 24 25 26 27 21 28 Turnover tax 2 0 0 0 0 0 0 0 0 0 Excise duties 8 6 6 6 6 6 6 6 8 7 Tax on international trade 25 21 21 23 21 21 20 20 17 15 Capital Revenue 1 1 1 1 0 0 1 2 1 0 Non-tax revenue 18 19 18 15 15 16 14 17 20 17 Grants 1 1 1 1 1 1 0 0 0 0 (percent of GDP) Tax Revenue 21 21 21 21 21 22 22 22 20 19 Total Revenue (incl. grants) 26 26 26 26 25 26 26 27 26 23 Total Revenue (excl. grants) 26 26 26 25 25 26 26 27 26 23 * Excludes privatization receipts. Sources: Fiji Islands Bureau of Statistics Key Statistics September 2000. IMF Staff Reports, various. Reserve Bank of Fiji Quarterly Report, various. Fiji Budget Papers 2002. 60 Table A.2 Kiribati Government Revenue 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 est (percent of total revenue and grants) Tax revenue 18 23 25 30 37 24 21 29 29 24 Income tax 5 7 7 9 12 9 7 10 11 8 Personal 3 3 3 6 8 5 4 6 5 4 Company 3 4 3 4 4 4 3 4 6 4 Taxongoods & services 1 2 2 2 3 2 1 2 2 0 Tax on international trade 11 14 16 18 22 13 13 17 17 16 Other tax 1 1 1 1 1 1 1 1 1 0 Non-tax revenue 29 40 43 48 28 52 53 49 48 48 Grants 54' 37 32 22 35 24 26 22 23 28 (percent of GNI) Tax Revenue 13 15 14 16 19 15 14 16 19 16 Total Revenue (incl. grants) 77 66 56 52 50 62 66 55 65 66 Total Revenue (excl. grants) 36 42 38 41 33 47 49 43 50 47 Source: IMF Staff Reports. various, Governrment Finance Statistics (GFS), November 2001 61 Table A.3 Samoa Government Revenue 93/94 94/95 95/96 96/97 97/98 98/99 99/00 00/01 (est) (percent of total revenue and grants) Tax revenue 63 50 48 57 59 55 62 64 Income tax 20 13 11 11 13 12 14 14 Tax on goods & services 17 19 20 21 25 30 35 37 VAGST 8 12 11 13 14 14 16 18 Excise duties 9 7 8 8 11 15 19 19 Tax on international trade 24 17 16 24 20 12 11 12 Import duties 23 16 15 23 19 11 11 12 Other 2 1 1 1 1 1 0 0 Othertax 2 1 1 1 2 2 1 1 Non-tax receipts 18 17 17 13 15 14 15 10 Grants 19 33 35 30 26 31 23 26 (percent of GDP) Tax Revenue 22 22 22 22 21 21 20 19 Total Revenue (incl. grants) 36 44 46 38 36 38 32 30 Total Revenue (excl. grants) 29 29 30 27 26 26 25 22 Sources: IMF Staff Report 99/28 1999, ADB Samoa 2000, Department of Treasury Budget Economic Statement various, Treasury estimates, Department of Statistics Annual Statistical Abstract various. 62 Table A.4 Tonga Government Revenue 92/93 93/94 94/95 95/96 96/97 97/98 98/99 99/00 00/01 (percent of total revenue and grants) Tax revenue 63 65 60 65 68 66 74 74 70 Income tax 10 11 11 14 15 14 17 15 14 Tax on income, profits & capital gains 10 10 10 14 15 14 17 15 13 Taxesonproperty 1 1 1 1 0 0 0 0 1 Tax on goods and services 9 8 10 10 9 10 10 11 11 Salestax 6 7 9 9 9 7 8 8 8 Othertax 2 1 1 1 0 2 2 2 3 Tax on international trade 44 46 39 41 44 43 47 48 46 Non-tax receipts 20 25 29 28 26 29 23 25 27 Grants 17 9 12 7 6 5 2 2 3 (percent of GDP) Tax Revenue 19 20 21 21 22 20 20 21 20 Total Revenue (incl. grants) 30 31 34 32 32 30 27 28 29 Total Revenue (excl. grants) .25 28 30 30 30 28 27 28 29 Sources: IMF-Statt Rep: iirious.'ADI3B Tonia Countrv ReDort 1996. Estimates provided by Government Statistics-EJ¶nint 63 Table A.5 Vanuatu Government Revenue 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 (percent of total revenue and grants) Tax revenue 73 74 81 72 79 83 84 81 85 84 Income tax 0 0 0 0 0 0 0 0 0 0 Tax on goods & services 21 25 24 29 31 32 38 46 48 48 VAT 0 0 0 0 0 0 12 31 35 34 Other 21 25 24 29 31 31 25 14 12 13 Tax on international trade 49 45 56 42 46 49 44 33 34 33 Import duties 47 44 54 40 45 47 43 33 34 33 Export duties 1 1 1 1 1 1 1 0 0 0 Other customs duties 0 0 1 0 0 0 0 0 0 0 Other tax 3 3 1 1 2 3 2 2 2 2 Non-tax receipts 19 19 14 17 14 7 11 11 10 10 Grants 8 7 5 11 7 10 5 7. 5 6 (percent of GDP) Tax Revenue 19 18 20 20 20 21 20 20 19 18 Total Revenue (incl. grants) 26 24 25 27 26 25 24 25 22 21 Total Revenue (excl. grants) 24 22 24 24 24 23 23 23 21 20 Sources: Government Statistic Office, Statistical Indicators first quarter 2001, Reserve Bank of Vanuatu, Quarterly Economic Review, various, Government of the Republic of Vanuatu, Budget 2001, Volume 1. 64 3. LABOR MARKETS AND PUBLIC SECTOR REFORM: EASING ADJUSTMENT TO TRADE LIBERALIZATION 1. Introduction 3.1 In the late 1990s a number of PICs embarked upon comprehensive reform programs while others are gearing up for the same. The triggers for these programs were actual or impending fiscal crises. The deterioration in fiscal positions were in turn the result of bloated public sectors, and weak economic, management exacerbated in some cases by unfavorable external events. The reforms generally entailed a rightsizing of the civil service, privatization or corporatization of SOEs, and improving governance. At the same time, steps were initiated for the creation of a conducive environment for the development of the private sector. 3.2 What implications do these public sector reforms have for the current efforts that PICs are undertaking to integrate their economies more closely with the global economy? It has been argued that a large public sector creates distortions in the economy: for example, by acting as a wage leader in the labor market it leads to wages that are well in excess of those justified by productivity levels. These distortions, if left unaddressed, could complicate the adjustment to trade reform, as a well-functioning labor market is a crucial factor in gaining a good response from economic reforms. Similarly, if the prices of infrastructure services provided by SOEs or poorly regulated private entities are high in relation to competitors, PIC producers will find it difficult to compete successfully in a liberalized environment. Thus, successful public sector reform that removes such distortions could facilitate the adjustment to trade reform. 3.3 However, opposition to public sector reform has been witnessed in some PICs, especially those downsizing. Such opposition could also generate or exacerbate resistance to reform in general, including trade liberalization.' While trade liberalization is unlikely to affect the poor and unskilled, the bulk of whom rely on subsistence agriculture for their livelihood,2 it will threaten relatively high-paying jobs in currently protected sectors and diminish the value of rents accruing to holders of import licenses. This would add to the labor displaced due to public sector reform, and the large cohort of working-age young people being added to the ranks of the unemployed every year. 3.4 The predicament for policy makers is that while trade liberalization has the potential to create additional jobs, experience of developing countries suggests a time lag between job loss and job creation. The length of this lag is dependent upon the ' Adjustment and hence opposition to PICTA is likely to be small. The discussion in this chapter should be viewed in the context of fuller trade liberalization which as argued in Chapter 1 may be a better strategy but would entail greater adjustment. 2Some 80 percent of the population in the PICs, the majority of whom have little education and marketable skills, live off the land and are likely to gain as a result of lower prices of imports and higher prices for traditional exports. In a similar light, Forteza and Rama (2002) note in the context of developing countries that only the urban middle class are adversely affected by liberalization and downsizing. 65 conduciveness of the domestic environment for investment and ease of labor mobility. Both of these can be addressed through elements of the reform program, but resistance to reform may build up during the time lag. Thus minimizing resistance through compensation mechanisms or other temporary income support programs may be warranted. The pace and composition of reforms may need to be calibrated to the pace of progress on efforts to enhance the business environment and increase labor and capital mobility. 3.5 To provide a backdrop against which public sector reforms and trade liberalization is to be undertaken, this chapter studies the characteristics of the labor market in PICs, especially the role played by the public sector. The wage leadership role of the public sector is strongest in the three Compact of Free Association countries, FSM, Marshall Islands, and Palau, as well as in Kiribati. Interventions in the form of minimum wage legislation do not appear to be an important distortion for resource allocation in most countries, as they are often circumvented. Other features of the labor markets include the shortage of skilled workers, an ever-growing excess supply of unskilled workers, and emigration, which acts as a safety valve but which exacerbates skill shortages. The chapter also briefly considers the impact of public sector involvement in other factor and product markets, and the manner in which this could inhibit private sector development. Some implications of public sector reforms in terms of their ability to ease the process of trade reform, mainly through encouraging private sector development, are discussed. The chapter outlines how resistance to reform may develop in the PICs and policies that could be adopted to address such resistance. It is argued that if downsizing is contemplated, severance packages could be offered to public sector employees to both diffuse resistance and ease their adjustment to reform. The workers typically offered severance packages are likely to be part of the urban middle class. Therefore, in addition to severance packages, income support programs like public works could be needed to assist people who might temporarily fall into poverty as a result of reform. The, job loss of an urban worker could adversely impact dependents who supplement their subsistence livelihood with small cash transfers from the worker. Income support programs could be helpful for these households. Other mechanisms to ease adjustment costs and reduce resistance to reform include retraining programs, enhancing labor mobility in the region, and improving the credibility of reforms. Implementing these programs to ease adjustment costs would place additional demands on PIC budgets in the short-run. In the past donor assistance has been forthcoming to support reform programs in the PICs. There is a sound case for such support to continue especially for reform programs that are strongly owned by the PIC govemments and stakeholders. 2. Factor and Product Markets in PICs: The Backdrop for Trade Liberalization 3.6 The PICs are small, have very limited trade with each other, and have a plethora of regulations governing the operation of domestic markets. The most pervasive of these regulations include: price controls on consumer goods; controls on workplace conditions; controls over investment; controls over expatriate labor; and, a relatively large state sector that acts as a leader of wages and working conditions in the formal markets. Land 66 is held under customary title involving communal ownership. Further, the absence of exclusive individual tenure renders land of little value in accessing credit from the formal sector. 3.7 The product markets are small, geographically fragmented, and isolated from major industrial countries, making the role of transactions costs in exchange non-trivial, while providing incumbents with significant market power. These considerations have been used to rationalize direct state participation in commerce, but this strategy has proved to be unsustainable, given the deteriorating fiscal position of the state-not least due to the poor profitability of the state sector. 2.1 Rapidly growing numbers of under-employed 3.8 Approximately two-fifths of the population in the PICs is below 15 years of age, implying that the population growth momentum is likely to be sustained for at least a decade and half (Table 3.1). The entry of large numbers of workers into the workforce in a climate of low growth of employment opportunities and wage rigidity will lead to a growing pool of under- and unemployed people. 3.9 Policy makers in the PICs have been concerned about the rising numbers of job seekers, particularly the young (Table 3.2). In Tonga and Vanuatu, for example, in any given year only 25-30 percent of school leavers can expect to find employment in the formal sector. The range would be similar if not lower for other PICs. There has been increasing awareness of the lack of job generation being a major source of social problems including crime. It is also a factor in the urban drift that is leading to the establishment of squatter settlements around major urban centers. To an extent, the recent civil unrest in the Solomon Islands highlights the potential danger of leaving such trends unchecked. The need to urgently increase employment on a sustainable basis in the PICs has been gaining recognition. 2.2 Role of emigration 3.10 Demand for jobs varies considerably between the PICs. Emigration of workers from Fiji, FSM, Marshall Islands, Palau, Samoa, and Tonga has acted as an escape valve for qualified workers in search of better employment opportunities. Citizens of FSM, Marshall Islands and Palau have emigration rights to the US under the Compact of Free Association. Samoa and Tonga have had reasonably easy access to Australia, New Zealand, and the US. An almost equal number of Tongans and Samoans reside abroad as live in these countries. The emigration opportunities have induced investment in skills acquisition. In Fiji, to facilitate emigration, the Indo-Fijian population is investing more in education than the indigenous population-the difference between the two ethnic groups was most apparent following the two coups of 1987.3 As a further illustration of this phenomenon, literacy rates in Palau, Samoa, and Tonga, where emigration is easy, 3 The economic factors underlying migration from the PICs have always been present but the recent surge in migration of Indo-Fijians can be attributed to 'push' factors. 67 Table 3.1: Basic economic and social indicators Country Population Young Per-capita Population Literacy Secondary Popln8 GNI growth rate Enrolment 1995-2000 Female Male ('000) (% of total) (2000 US$) (% per annum) (adult, %) (%) (%) Fiji 810 33.3 2160 1.1 93 35 37 FSM 118 43.5 2110 1.9 72 43 45 Kiribati 91 41.2 950 2.3 93 47 42 Marshall Islands 52 43.0 2210 1.5 97 51 47 Palau 19 27.0 7120 1.8 92 na Na Samoa 170 41.2 1420 0.8 96 75 66 Solomon Islands 430 44.8 600 2.8 30 18 30 Tonga 100 37.9 1630 0.6 99 71 64 Vanuatu 197 42.0 1100 2.2 34 18 26 a) Young population is defined as those in the 0 to 14-year age group. Sources: ADB (2001a), Bank staff estimates, and http://www.photius.com/wfb2000/countries/palau/palau_people.html. Table 3.2: Annual supply-demand imbalance for formal sector employment Country Total formal Formals sector Fresh entrants sector share in work-age into the workforce employmente population per year ('000) (%) ('000) Fiji 131.5 26.6 19.3 FSM 30.9 49.1 3.7 Kiribati 9.2 19.7 1.8 Marshall Islands 10.3 37.4 1.6 Palau na na 0.4 Samoa 23.0 na 5.0 Solomon Islands 34.2 14.6 14.7 Tonga 33.9 59.5 2.7 Vanuatu 14.3 14.4 5.6 Note: a data is for 2000 except for Solomon Islands and Tonga where the most recent data available, that for 1996, has been reported. Work-age population defined as those in the 15 to 64 year age cohort. Data for column 3 are calculated assuming a uniform distribution of the population of under-14s. Source: ADB, 2001a and Bank staff estimates. 68 are high as compared to the Solomon Islands and Vanuatu where emigration to high- income countries is difficult.4 3.11 Easy access to emigration and the remittances sent back by migrants to their home country have two effects that push up domestic wage rates. First, the wage rate is influenced by the higher level of wages that a potential emigrant can hope to earn in the high-income countries. Second, the remittances sent back by the emigrants provide cash to supplement an otherwise subsistence livelihood. This tends to raise the reservation wage that would induce a subsistence person to engage in wage labor. 3.12 In the short to medium term, the ease of emigration can reduce the adverse effects of demand shocks on employment in the PICs. However, as noted below, the sustained loss of skills through migration can reduce employment opportunities for the unskilled, given the complementarity between skilled and unskilled labor. 2.3 Shortage of Skilled Workers 3.13 Two stylized facts about labor markets in the PICs are: first, a persistent excess demand for professional, technical, and managerial workers, even though these are not large markets. Second, there is an excess supply of unskilled workers. For example, the tourism and garment sectors in Fiji experience excess supplies of labor with basic skill levels but have problems recruiting workers with higher skill levels. The thin market for skilled workers means that the price (and excess demand) for their labor is likely to be more volatile than otherwise. For instance, the departure of even a single doctor from Tarawa (Kiribati) severely limits access to medical services. Emigration of skilled workers aggravates this volatility problem. Wages of many professionals rose sharply in the aftermath of the 1987 coups in Fiji that saw high rates of emigration of people with such skills. In the short- to medium-run the only means of reducing this volatility is by facilitating access to expatriate labor, but this entails high transaction costs. All PICs limit expatriate work permits in selected skill categories. Application procedures for these work permits tend to be time-consuming, and they are often revoked. 3.14 A related problem is the lack of a critical mass of skilled workers needed to facilitate development of businesses which could then draw on the large pool of less- skilled workers, i.e., there is a complementarity between skilled and unskilled labor. The wire-harness factory in Samoa is a case in point where production is confined to a highly specialized output; expansion both in the quantity of output and span of the product range is constrained by the lack of a sufficiently broad and deep skill base. 3.15 A further issue with thin markets, particularly when there are biases against employment of expatriates, is the mismatching of jobs with incumbents. It is common in the PICs to have senior positions, particularly in the state sector, filled by poorly qualified candidates. This problem may be compounded by the poor retention rates of post- 4 While higher education facilitates emigration from the PICs to Australia and New Zealand, for example, the availability of such an option (for whatever reason) provides the incentives and funds for greater investment in education. 69 secondary trainees who, once trained, have the option to emigrate to higher paying jobs in the neighboring high income countries (see ADB, 1998). 2.4 Public and Private Sectors 3.16 The PICs are characterized by a relatively large public sector that accounts for a large share of formal sector employment.5 An indication of the relative size of the public sector is provided in Figures 3. land 3.2. The importance of the wage bill in government Figure 3.1: Share of Public Sector Employment in Paid Employment 80% 70%/o 60% 50%1/ 40%/ 30% 20% 100%- 0% *The figure was nearly 50% in the mid-nineties. Figure 3.2: Government Expenditure (in % GDP) 120% 100% 80% 60% 40% 20% 0% 5 Due to the fixed costs of providing public 'services, public sectors in small economies are likely to be proportionally larger than in larger ones. The point here is that public sectors are large even after controlling for lack of economies of scale. 70 expenditure is highlighted in Figure 3.3. Government fiscal deficits, of which these wage costs are an important cause, are presented in Figure 3.4. Generous inflows of official aid have in most cases helped sustain the large public sectors. The state has been involved in a wide range of activities such as utilities, development finance, marketing of agricultural products, operating fishing fleets, mines, plantations, aviation, and hotels. Figure 3.3: Wage Bill as a Share of Government Current Expenditure 60% OAve. 95-96 50% -Ave. 99-00 40% 30% 20% 10% 0% CO cu CO co 5%~~~~~~~~~~ - .m (0 E E a 0 ~~~~0> Figure 3.4: Government Fiscal Deficit (in % GDP) 15% - 10% - E~~~~~~~~~~Ave. 1996-98 * Ave. 1999-01 5% 0% - -10% -15% - -20% - 3.17 The private sector has remained very small. In some countries, the growth that has occurred in the private sector has been largely to service the large public sector. Or where sufficient protection has been provided, import-competing firms have been 71 established. There has also been private sector growth associated with the development of natural resources such as minerals, logging, fishing, and tourism. Some industries, including sugar production, garment manufacturing, and fish canning have developed under preferential access to industrial country markets (see chapter 1). 3.18 Since the public sector has had access to aid funds and resource rents, it has historically been able to attract workers away from the private sector by paying higher wages (see Duncan et al,1999). The high wages paid by the public sector make labor- intensive production by the private sector uncompetitive, except when protection is provided. However, private sector activities that have to rely upon the restriction of imports will never be competitive internationally. Their opportunities for employment will be limited by the small size of the domestic market. Wage leadership by the public sector is clearest and strongest in FSM, Marshall Islands, Kiribati, and Palau. In FSM and Marshall Islands, public sector wages are in many instances double those paid for comparable skills in the private sector. In Tonga, there are instances of the private sector paying more that the public sector, especially at the senior levels. However, even in cases where the private sector pays higher wages, until recently, workers chose the public sector due to its attractive pension scheme. In Solomon Islands, wages in the private sector are two-thirds indexed to inflation, but in the public sector they are fully indexed. 3.19 Wage leadership by the public sector has led to a bias against labor-intensive sectors such as agriculture and tourism, which are the exportable sectors. Hence, the internal terms of trade, one measure of the real exchange rate, has been biased against and, therefore, detrimental to exports. The high wages prevailing in the state and importable sectors-sectors that are always concentrated in the urban areas-have created a rural-urban dichotomy as well as a formal, unionized sector versus an informal, unskilled and non-unionized worker divide. 3.20 The dichotomy between unionized and non-unionized workers is most apparent in Fiji. Attempts to deregulate the labor market have met the greatest resistance where senior civil servants and union officials have contested and won elections on the platform of 'protection of worker rights'. Workers in the informal and non-unionized sectors have paid a price in terms of lower wages, but their continued silence about the issue may be attributed to the dispersed interests of the masses and, perhaps, a lack of information. 3.21 In the state sector and the highly-protected importable sector, the linkage of wages growth to 'cost of living adjustments', qualifications and years of service, and seniority of positions has de-linked wages from productivity and removed incentives for investment in skills upgrading. In the process, the wage structure has been compressed across the public sector and indeed across the entire wage sector.6 Collectively, these factors have made productivity-adjusted wages in the PICs uncompetitive with those in competitor countries. 6 Demekas and Kontolemis (1999) note that public servants accept wage compression due to the high non- wage benefits. 72 3.22 Since pay scales rise with seniority in the administrative hierarchy, professionals such as surgeons often seek promotion into senior administrative positions. While this brings them higher salaries, it is at a cost to society in terms of the loss of their much- valued skills (see ADB, 2001 in relation to employment of surgeons in Vanuatu). 2.5 Minimum Wages 3.23 Conditions of employment differ considerably across the PICs; legislated minimum wages7, number of paid holidays, and compulsory superannuation contributions by employers add to the costs of hiring workers (Table 3.3). Table 3.3: Labor conditions in PICs Country Legislated Average Length Paid holidays Compulsory minimum wage of work week superannuation contribution ($US/hr) (hours) (days/year) (% of wage-bill) Fiji 0.63-1.09 45 13 8 FSM 1.35 - - Kiribati 1.08a 36 13 6 Marshall Islands 2.00 Palau 2.50 Samoa 0.45 40 15 5 Solomon Islands 0.30 40 11 7.5 Tonga - 40 11 nib Vanuatu 0.71 40 14 6 Notes: a. Not legislated. b. Under a new scheme there will be a 5% contribution rising to 10% in 2010. Source: Forum Secretariat: Foreign Investment Climate in Forum Island Countries, June 2001 and Bank Staff Consultations. 3.24 Employment is also subject to regulations such as workplace minimum standards of employment, leave and redundancy entitlements, and unfair dismissal laws. These regulations, put in place to safeguard workers' rights, tend to raise the costs of hiring workers and could depress employment. Enforcement of these regulations, particularly of minimum wages, varies considerably across countries and between sectors. Where the statutory minimum wages are much higher than what the market will bear, collusion between workers and firms takes place to circumvent the regulations. Kiribati, Samoa, Solomon Islands, Tonga and Vanuatu have minimum wage legislation that is not strictly enforced. There are also instances when minimum wage rules have been relaxed by the government to encourage foreign investment, for instance, for a fish loining plant in the Marshall Islands. By comparison, Fiji has an elaborate system of Wage Councils, eight 7As a comparison, unskilled wages in China, India, Indonesia and Thailand are 0.25 to 0.50 cents per hour. 73 in total, that monitor and enforce compliance with minimum wage legislation and workplace conditions at the sectoral level. Unions are a major force in wage negotiations and in politics in Fiji. But even in Fiji there is evidence that minimum wages are often circumvented. 3.25 Weak enforcement follows when costs of regulation are high. This is an important reason for poor enforcement of regulations in the rural sector where state presence is low. Consequently, the rural and informal sectors operate with a lot more flexibility than the formal sectors; but the flexibility is not able to offset the adverse impact of the over-valued real exchange rate. The wage "premium" in the formal sector reflect the "rents" accruing to "insiders" as a result of the protection afforded. As mentioned earlier, the higher wages in the urban sector also draws workers out of the rural districts in search of more highly-paid employment. 3. Rigidities in other markets 3.1 Land 3.26 Most of the land in the PICs is communally owned. In Fiji, Solomon Islands and Samoa nearly 85 percent of the land is under communal title. Nearly all of the industrial developments and much of commercial agriculture have been confined to alienated or state-held land. Market gardening and subsistence agriculture, activities that occupy nearly 80 percent of the population have, as a consequence, relied on communally-owned land. Customary forms of land ownership and kinship traditions of sharing resources have served as an informal social security mechanism and have helped the PICs in avoiding absolute poverty. 3.27 However, the contribution of agriculture of around 20 percent of GDP is low relative to the share of the labor force engaged in the sector and is reflective of the low productivity in the sector. With some 80 percent of the population accounting for 20 percent of GDP, the income ratio between agricultural households and the rest of the population is 1 to 16. Even if the poor quality of production data is accounted for, the disparities in income between rural households dependent on agriculture and urban formal sector workers are still large. These disparities may be narrowed through an increase in agricultural productivity, a fall in protection provided to the importable sectors, and freeing up of the factor markets. 3.28 Agricultural development in much of the developing world is almost uniformly related to the emergence of rural credit markets. Chalamwong and Feder (1988) note that secure title to land assists in access to competitively-priced credit to the owners. The authors note that: "a clear formal title backed by a legal system capable of enforcing property rights reduces or eliminates the uncertainty regarding tenure" (p. 190); and consider the ability to use such land as collateral to access credit as a major benefit to farmers. Access to credit facilitates investment and thereby raises productivity in the sector. Even in urban areas in the PICs the private sector has difficulty in acquiring land 74 title, which in turn discourages investment in fixed plant, equipment, and buildings that are immobile and have long economic life spans. 3.2 Retail markets 3.29 Administrative mechanisms have been employed in some of the PICs to control prices of final goods; these impede market clearing. For example, some 50 percent of final goods and services, including basic food items and some rental leases, are under price control in Fiji. Prices are controlled by means of a ceiling on mark-ups allowed on the costs of inputs; such a formula provides no incentives for productivity enhancements. In competitive sectors such as retailing, this places a ceiling on prices charged that bind during periods of demand surges. Import licenses have been occasionally used to maintain domestic price levels; the case of rice in Fiji is a recent example. The tariffication of such quantitative controls and gradual reductions in the tariffs as agreed under PICTA will facilitate competition, but in the process it will induce adjustments to factor returns and income. 3.30 The relatively small size of PIC markets limits the number of suppliers; this often confers market power on the lead player.8 Moreover, natural monopolies such as the utility providers face costs higher than their counterparts in larger markets; some of these costs are reflected in the higher price for these services. The presence of market power in an environment of limited public sector capacity to enforce, competitive pricing leads to inflated costs for business inputs. State dominance in several of these sectors, including in international transportation and utilities, adds to the cost pressures for two related reasons. First, there is a clear conflict of interest in the state acting as a regulator of commerce while being actively engaged in supply. Second, budgetary support to SOEs and political interference in their operation leave little incentive for productivity improvements. As a result, utility and transportation costs in the PICs are typically high, even after controlling for lack of scale economies (Table 3.4). 3.31 Regulatory barriers to the operation of the informal sector in several PICs have constrained its growth at the cost of generating high involuntary under-employment together with lost production. The health and safety standards in the PICs are 'borrowed' from Australia and New Zealand. The inflated costs of licensing and meeting quarantine and health requirements have constrained the growth of the informal sector. This in turn has limited employment generation in the informal sector. The traditional subsistence sector has acted as the "sink" for under-employed labor. This may have acted as a drag on growth given developing country experience that suggests a transition to the formal economy from subsistence agriculture via a healthy and competitive informal sector (Duncan et al, 1999). 8 This is akin to what Harberger (1988) calls the tyranny of demography. 75 Table 3.4: Utility costs in PICs Country Electricity' Telephone Cal to USb (US$/kWh) (US$/minute) Fiji 0.16 1.10 FSM .10-.16 2.00 Kiribati 0.24 2.10 Marshall Islands 0.16 2.00 Palau 0.10 1.50 Samoa 0.18 1.35 Solomon Islands 0.16 1.10 Tonga 0.31 1.20 Vanuatu 0.26 1.50 Australia -0.06 -0.17 Notes: a. Average commercial tariff 1999. b. Representative rate; converted to US$ using the average exchange rate for 2001. Source: "Regulatory framework and transaction models for private participation in infrastructure in Pacific Island countries." Draft report prepared for the World Bank by Meritec Ltd, New Zealand, and Bank staff consultations. 4. Public Sector Reform 3.32 Many PICs embarked upon comprehensive reforms in the mid to late-1990s. Triggers for these reforms were provided by deteriorating fiscal positions, which were the result of weak economic management and, in some cases, the need to bail out failing SOEs. Economic management was also complicated by external events such as the Asian financial crisis. In the case of the US Compact countries, the urgency for reform arose from the need to significantly reduce public expenditure in the face of declining foreign aid from the US. The reforms generally entailed a rightsizing of the civil service, privatization or corporatization of SOEs, and improving governance. At the same time, steps were initiated to create a conducive environment for the development of the private sector. 3.33 Public sector reforms have not met with the expected success. According to a recent review by the ADB (Knapman and Saldanha, 1999), the reforms faltered on two counts. First, the priority given to fiscal stabilization led to a focus on reducing public expenditure through downsizing at the expense of public sector effectiveness. Secondly, sufficient attention was not given to addressing the resistance to change that is generally exhibited by institutions and people in the face of reform. Doing so would have entailed giving them more time to acquire new skills and integrate new values and motivations.. 76 3.34 What impacts are public sector reforms likely to have? In the best case, public sector reform would tend to improve the flexibility of the labor market, reduce the cost of public utilities, and by creating the right conditions for the private sector to develop, provide employment opportunities for the displaced labor. The best case may, however, not materialize. The above-quoted review of the reform programs also noted that insufficient attention had been paid to the capacity of the private sector in PICs to provide the assumed response to public sector downsizing, and the liberalization of policy frameworks. 3.35 Reforms undertaken so far have not addressed the minimum wage issue. However, as noted earlier, minimum wages have perhaps not had a large impact on employment because they have been circumvented. Thus, even if the issue had been addressed, the impact on employment would have been small. 3.36 For PICs with good opportunities for emigration, it has been argued that a reduction in the public sector work force may not lead to a decline in the wage rate. This is because the wage rate in these countries is influenced by the wage which emigrant workers are able to earn in the host country. A downsizing of the public sector could lead to heavy emigration as the experience of Cook Islands, and, to a smaller extent, Marshall Islands, demonstrates. To the extent that the private sector had developed to service the public sector in these countries, the private sector could then be expected to contract, especially if those who emigrate are the more highly-skilled people. 3.37 The role of the public sector as a redistributive mechanism for aid and resource rents is often put forward. It is argued that as long as aid flows and resource rents can be expected to remain large, there is no need to downsize the public sector. However, the maintenance of a large public or non-tradeable sector is at the expense of tradeable sectors. Another negative aspect of this form of development is that since the non- tradeable sector is located mostly in urban areas, the increase of population in urban areas would have attendant environmental and social problems. 3.38 Once protection is lowered, attempts to maintain the high wages will surely render the import-competing industries uncompetitive. The same is true of some export industries once preferential access or other subsidies are removed. But even if the protected wages fall, the private sector will not automatically develop to absorb the displaced labor. It is here that the aspect of public sector reform that aims to create an enabling environment for the private sector assumes great importance. If the right environment for the private sector to develop is not available, and emigration possibilities are limited, the displaced labor will either retreat to subsistence or remain un-or under- employed. 41 Resistance to Reformn 3.39 Opposition to reform is likely to come from those within the public sector who occupy privileged positions, earning wages in excess of their productivity, but facing job loss. In a climate of rising (youth) unemployment, these job losses create fertile grounds 77 for mobilization of opposition to the reforms. In countries where governments are formed by weak coalitions with a slender majority their very retention of power could be threatened. Realizing the predicament, political leaders pushing for reforms have lost some of their enthusiasm, as seen in Fiji, Solomon Islands and Vanuatu, for instance, attesting to the political difficulties of implementing and sustaining reforms in the PICs. While the opposition to public sector reforms comes mostly from the public sector, it could easily initiate or exacerbate opposition to any reform, including trade reform, that threatens privileged or protected jobs. 4.2 Private and Social Costs of Reformn 3.40 In understanding how opposition to reforms may develop it is useful to distinguish between the private and social costs and benefits that adjustment to policy reform may entail. Social costs and benefits are related to the impact of reforms in an aggregate or economy-wide welfare sense. Private costs or benefits are born by individuals or groups of individuals with similar concerns. It is the distribution of private costs that needs to be considered to examine how opposition develops. Even when the social benefits of trade liberalization outweigh the social costs, the existence of private costs can generate enough political opposition to block the reform. Consider the case where the reform involves removal of protective tariffs for an industry. The potential gains (the reduction in price that this would presumably lead to) could be large in the aggregate, but they could be spread over a large number of consumers. Thus the potential gains to each individual would be quite small, whereas the loss to a few producers could be substantial, to the point where they may be forced to close down. 3.41 Given their dispersed nature and large number, consumers would find it difficult and costly (relative to their individual gain) to lobby in favor of the reform. There is also the free rider problem. Individual consumers do not have an incentive to lobby for the reform, but would like others to do so, as they stand to gain if the reform is implemented. With each individual thinking in a similar fashion no lobbying takes place. They remain the silent majority. The vocal minority (the owners of capital and the labor employed in the industry) that opposes the reform is a more cohesive group and faces costs of lobbying that are small relative to the negative impact of the reform. They will then be successful in stopping the reform. 3.42 It is also possible that even those who are likely to gain from reforms may, in fact, oppose them. Some workers who possess the requisite skills may be able to make a smooth transition to higher wage employment in an exporting sector from a previously protected sector. It is likely that a good proportion of the workers would make this transition over time, but it is impossible to precisely identify the actual individuals who would benefit. Thus it is quite possible that workers in protected industries would attach a higher weight to the downside risk of not finding a job than to the upside potential of alternative employment. 78 5. Reducing Resistance to Reform and Easing Adjustment 5.1 The Economic Argument 3.43 To gauge whether public sector reforms can be successfully implemented along with trade liberalization, it is useful to distinguish between the economic and political arguments. The economic argument for reform focuses on distortions created by public sector wage policies through payment of wages in excess of productivity, which tends to raise wages paid by. the private sector, or through intervention in the labor market in the form of minimum wage legislation or mandatory benefits. The reform policies adopted may include freezing or abolishing the minimum wages and mandatory benefits. Such reform could be expected to lead to downward pressure on wages which would ease labor reallocation from the contracting to the expanding sectors. It was argued above that in most PICs minimum wages do not appear to have had an adverse impact on employment and so reducing them directly may not be the best policy. But even where they have been binding, a better policy than getting rid of them may be to let their real value decline over time by not fully adjusting them for inflation.9 The economic argument also leads to the conclusion that in many countries placing greater emphasis on private sector development by addressing impediments to private investment may be more effective in increasing employment than focusing so heavily on reducing public sector activity. 3.44 Improving the environment for the private sector would involve, among other things, improving the policy regime for attracting FDI (see Box 3.1) and finding a way to address land issues in a manner that is culturally acceptable to the PICs. Providing an encouraging environment for foreign investors to establish joint ventures will be crucial for the development of export industries in liberalizing PICs given that domestic firms have had little or no experience in exporting. Ensuring the credibility of reforms and enhanced labor mobility will be other components of a program to encourage business development. 3.45 The key issue for land may be not one of changing the land tenure system per se, but of unlocking land tied up in unproductive uses because of inflexible rules governing its ownership and use. This involves improving land leasing mechanisms through greater community participation (World Bank, 1998). It is also interesting to note the following from an authoritative work on the Pacific (Hughes 1998 pp. 105-106): "...evidence from land tenure studies in all three cultural zones of the Pacific has shown how the ways in which people occupy and use land, and the bundle of use rights they are able to assert vis- a-vis other families and individuals have changed under pressure of introduced crops, technologies, more people and monetization. Enterprising persons have developed balancing skills that enable them to respect custom for its unifying, identity-strengthening properties while modifying its details to accommodate individual investment, accumulation and distribution of wealth on a more or less permanent basis: a kind of informal perpetual leasehold, where tenure is uninterrupted provided reasonable obligations are fulfilled." 9 Removing the minimum wage may be politically more difficult than letting its real value erode over time. 79 Box 3.1: Increasing FDI inflow into the Pacific: Some Generic Advice The environment for foreign direct investment (FDI) which a country offers can be considered as a product which the country is trying to sell to foreign investors. So, making more sales will generally be dependent on improving the product and/or marketing it more effectively. The investment environment - the product - is an amalgam of a large number of factors, including natural resources, infrastructure, regulatory policies, and governance. These can be grouped in a number of ways: for example, those that the government can affect and those that it cannot. One factor which is of relevance to PIC governments (particularly those that face financial resource constraints) is the cost of making changes. While building better transport infrastructure or improving educational levels are generally costly in terms of public sector expenditure, changing policies (that is, legislation and regulations) and procedures can be done with minimal budgetary cost. For this reason, reform of the policy and procedural environment for the entry and operation of FDI requires only understanding by governments and other stakeholders of the importance of these aspects to the overall investment environment and the political will to carry through the necessary reforms. There are some fundamental issues that relate to sound FDI policy. Irrespective of the level of access which the government considers foreign investors should have to the local economy, simplicity, transparency and the minimization of discretionary powers for officials and ministers are key features of good legislation and regulations. Similarly, transparency and predictability are important elements of the procedures which are used to implement the formal policies, with the availability in the public domain of these procedures - and the criteria which are used with them - an important positive factor. It is important to understand that transparency and predictability are not synonyms for "liberal" and "open". While economic arguments can be made for the potential benefits which will accrue from more liberal access for FDI, the extent of the access provided to foreign investors is a political choice for the government concerned. Deciding to maximize transparency and simplicity, however, while still requiring a choice by governments, is more a matter of exercising common sense to fully benefit from any policy position that might be adopted. One aspect of investment policy which has dominated much of the thinking about FDI in the Pacific in recent years is the question of incentives and the extent to which a country's incentives are competitive with those offered by its regional competitors. Incentives-based competition for FDI is financially destructive and leads to a "race to the bottom". There are investor-friendly alternatives to tax holidays (which have been the main incentive offered in most countries), e.g. accelerated depreciation and loss carry forward, which are less damaging to government revenues and which exert less distortionary impact on the economy. Any incentives offered should be performance-based, rather than expectation- based, and open to all investors who meet defined criteria, that is, the discretionary award of incentives should not be a part of the process. Marketing the "product" is normally done through the establishment of some form of investment promotion agency (IPA). For maximum effectiveness, an IPA should not have a regulatory role, instead being wholly dedicated to promotional activities. While there are a range of functions which an IPA can take on, the most important in a small country with limited budgetary resources is "investor servicing". An IPA which provides assistance to potential investors from the time of their first enquiry through to the point at which they start production (and even after that point, maintaining effective contact) can greatly increase the percentage of potential investors who are converted to actual investors (and, with appropriate "aftercare," become re-investors). Source: Foreign Investment Advisory Services (FIAS), World Bank Group. 80 5.2 The Political Argument - Severance Packages 3.46 The political argument focuses on opposition to reform which may either stop the process or lead to only a half-hearted implementation. From this perspective, the appropriate complement to adjustment programs would not be removal of regulatory distortions in the labor market, but rather -the introduction of mechanisms that compensate the workers affected by the reforms, even at least partially. Examples include job separation packages, early retirement programs and unemployment benefits. Some PICs have used severance packages as part of their programs of public sector downsizing. There have, however, been weaknesses in their design, and, importantly, on the pace of the downsizing. These two issues will remain important. The pace of downsizing could be calibrated to progress on creating a conducive environment for private sector development so that other employment opportunities are available. Importantly, the impact of downsizing on public service delivery should be addressed before the decision is made. 3.47 If downsizing is contemplated, research done at the World Bankl0 suggests that a menu of alternatives could be offered. For instance, civil servants could be given the following choices: (a) keep their job, (b) leave and receive severance pay, or (c) keep their job, but with a. fixed-term contract at a higher salary. This last option would help retain the more productive civil servants who have good outside alternatives and are not afraid of losing their jobs. Without option "c", those employees would take the severance pay and leave. 3.48 The experience of developing countries suggests that there may be a tendency to overcompensate by offering "golden handshakes." Over a sample of 41 downsizing programs financially supported by the World Bank, Haltiwanger and Singh (1999) estimated the average spending per job separation at $2,400. The figure was as high as $13,000 in the civil service of Senegal, $16,000 in,the mining sector of Bolivia, and $17,000 in the public enterprises of India, although all three countries had a per capita income of less than $1,000 per year. 3.49 The downsizing decision should not be based on a financial calculation alone e.g. only the impact on the wage bill, but should try to estimate the economic returns (impact on aggregate output and welfare) of the program. The situation in some PICs is in certain respects similar to "one-company towns" where public sector downsizing leads to a contraction of the private sector. Here financial returns of downsizing could easily be less than the economic returns. In these cases, if downsizing is considered,. it has been suggested that compensation should not be restricted to the retrenched workers, but could be extended to the community as a whole (Rama and Scott, 1999).1" In other instances, a reduction in the wage bill could be accompanied by a deterioration in public service delivery, as has occurred in some PICs. Thus a financial saving may lead to a decline in overall welfare. 0 See http://www.worldbank.org/research/projects/downsize/index.htm " Or, as argued above, greater emphasis should be placed on improving the environment for private sector businesses to develop. 81 3.50 Severance packages differ from safety nets that are designed to prevent households from falling into poverty as a result of the reforms. Examples include public works programs and means-tested income transfers. However, the workers typically offered severance packages are better seen as part of the urban middle class. Adjustment might dramatically reduce their living standards, but relatively few of them become poor, even several years after losing their jobs (see Rama, 1999). As a result, these workers may not be eligible for, or not interested in, most of the safety net initiatives. By recommending compensation the political argument would aim at mitigating (or even offsetting) the income loss of the urban middle class. Even though these workers are not poor it can be socially efficient to provide a one-time compensation in order for reforms to proceed smoothly. It would be expected that such workers would then not re-enter the public sector. But in the PICs there have been examples where workers made redundant have found their way back in. This could be either because the workers who were made redundant were in fact valuable in the service and should not have been let go in the first place. Or it could be that redundant workers have been re-hired to gain (or regain) political support. 3.51 The affordability of such programs would be an issue for PIC government budgets. However, a sound case can be made for financial assistance from donors to support such compensation packages under well designed, strongly owned and credible reform programs. 5.3 Public Works Programs 3.52 In the near absence of formal safety nets in the PICs, reliance on an extended family system has served as an informal safety net. But the rising number of un- and under-employed is placing undue pressure on traditional support mechanisms. There is anecdotal evidence to suggest an increasing number of dependents per formal worker, particularly in the rapidly growing urban centers of the PICs. Loss of formal sector work arising out of reforms will adversely affect all such dependents, particularly those relying on a small cash-income transfer from a wage eamer to supplement an otherwise subsistence livelihood. If the wage earner loses the job because of reforms, an income support program would prove beneficial to the dependents. 3.53 One income support program that is designed to reach the poorest is public works. Public works programs are typically used for building infrastructure such as village water tanks and rural roads. The work can be physically demanding. The programs provide low-wage jobs for all those who are willing to take them. The wage rate is a very important part of the program. Typically, it is set at a rate lower than that for unskilled labor. The low wage will attract only those neediest workers who do not have an alternative employment opportunity. The low wage also ensures that people have an incentive to move to regular work at market rates when it becomes available. It is important that the projects selected are technically and economically feasible so that costs can be recovered. It would be best if the projects are selected by local community groups. Since costs are to be borne by govemments it is important to ensure that a large burden is not placed on the budget. Alternatively, well-designed public works programs 82 with technically and economically feasible projects could attract donor support. This is especially so when the programs are also seen as a part of an overall reform strategy of the government that includes protecting those adversely impacted by the reforms. 5.4 Training or Retraining Programs 3.54 Active labor market programs are often set up to facilitate mobility between jobs lost and those created in the process of reform. These include training, counseling, placement services, and assistance in job search. Experience from other countries suggests that these programs have been popular with governments, trade unions and the general public, but the payoffs are uncertain or limited. Better information is required than is currently available in order for the PICs to have a good grasp of this issue and its application to their particular circumstances. While these programs are funded by the government it may be better to allow the private sector to implement them. An important issue in the PICs is the absence of private sector providers of such services. Another concern is a lack of information on the types of skills that are in demand. A possible option could be to provide state-funded apprenticeships in private firms. This would also, to some extent, ensure that the acquisition of skills is demand-driven. The provision of training could also be tendered out to educational institutions such as the University of the South Pacific and other regional institutions. One area where training programs could be helpful is in the development of skills required to start-up and operate micro and village enterprises. 3.55 Programs designed to retrain retrenched workers to make them more employable may reduce the adjustment costs if they have the desired effect of shortening spells of unemployment. Training programs providing income allowances to trainees can also be viewed as income support programs, similar to public works programs. Funds for retraining are provided by the government, generating a 'product' in the form of improved skills. Also, since participation in a training program is not likely to be compatible with holding down a job, there is an element of self selection of the beneficiaries. As with other programs for easing adjustment costs of strongly owned reform initiatives there is a good case for donor support for training programs. 5.5 Labor Mobility 3.56 In countries where skills are lost due to emigration or where wages do not decline because of easy access to emigration, restricting emigration should not be the favored remedy. Instead, attention should focus on making employment of skilled expatriates easier. This policy would be helpful for all PICs and could be done in the context of increasing labor mobility in the region. Increased labor mobility between countries would also be beneficial for the current move towards regional integration as this would make it possible to widen the integration to include services. In chapter 1 it was noted that extending the PICTA to cover services is likely to increase its benefits. Air transport and commercial shipping appear to be twG obvious focal areas for the PICs. One aspect of service liberalization is that service providers may need to be established locally and this may involve labor mobility to establish temporary or even permanent residence. It 83 would be worthwhile to begin a debate on the issue of labor mobility among the PICs, Australia and New Zealand. 5.6 Credibility of Reforms 3.57 Policy stability and credibility of reforms are critically important in lowering adjustment costs and lowering resistance to reforms. Policies that are announced in advance and adhered to are likely to induce investments in anticipation of the benefits from such investments. In contrast, frequent and unpredictable changes in policies create uncertainty and raise the costs of investment. 'Backsliding' on reforms generates expectations that further resistance to reforms will ensue. Credible policies create expectations that lower transition costs. As noted in chapter 1, credibility of reforms may be increased if they are locked-in with a regional or multilateral agreement. It was also For this mechanism to be effective, a high-income partner should be ready to reward good policies, but also, importantly, to impose sanctions in the event of backsliding. While PICTA may not be able to serve as the lock-in mechanism, this role could be played by the EU, Australia and New Zealand in RTAs with the PICs.'2 6. Conclusion 3.58 Faced with a changing global trade and aid environment, PICs have entered into a free trade agreement (PICTA) as a stepping stone to more closely integrating themselves with the world economy. In chapter 1 it was argued that widening liberalization beyond PICTA would potentially bring greater benefits. In order to exploit the potential benefits of wider liberalization, complementary reforms will be required in other sectors of the economy. Chapter 2 examined reforms that may be needed in the taxation area, while this chapter studied the interaction of public sector reform with trade liberalization. Key to the successful implementation of these reforms is a conducive environment for private businesses to develop, and greater labor and capital mobility in the region. The pace and composition of trade liberalization and public sector reform may, therefore, need to be adjusted and calibrated to the pace of progress on efforts to enhance the business environment and increase labor and capital mobility. The composition and pace in different PICs -should reflect their specific circumstances - which, as the discussion in the report has shown, can be quite varied. In order to ease the adjustment that reforms entail, govemments can adopt policies to compensate those impacted. In the past, PICs have had considerable support from their development partners in designing, financing, and implementing their reform programs. This support should continue especially where domestic ownership of the reform program is strong. The report has highlighted areas where such assistance would be crucial as PICs enhance their efforts towards greater global integration. 12 This role could be a key role played by the Cotonou and PACER agreements. 84 References Asian Development Bank, 1996. 'Tonga: Economic Performance and Selected Development Issues". Manila. Asian Development Bank 1998. "Improving Growth Prospects in the Pacific," Pacific Studies Series, Manila. Asian Development Bank, 2000. "Samoa 2000: Building on Recent Reforms," Manila. Asian Development Bank 2001a. Growth and Change in Asia and the Pacific: Key Indicators, Manila. Asian Development Bank 2001b. "Vanuatu 2001: Strengthening Development Policies," Pacific Islands Economic Report (PIER) Series, (draft), Manila. Asian Development Bank 2001c. "Republic of Marshall Islands, Meto2000: Economic Report and Statement of Development Strategies," Manila. Bagwell, Kyle and Robert W Staiger 2001. 'The WTO as a Mechanism for Securing Market Access Property Rights: Implications for Global Labour and Environmental Issues," Journal of Economic Perspectives 15(3): 69-88. Becker, Gary S.; Kevin M Murphy; and Robert Tamura 1990. "Human Capital, Fertility, and Economic Growth" Journal-of-Political-Economy 98(5 S 12-37. Centre for International Economics (CE), 1998. Costs and Benefits of a Free Trade Area between Forum Island Countries and Australia and New Zealand, Report prepared for AusAID and the Government of New Zealand, CE., Canberra and Sydney. Chalamwong, Yongyuth and Gershon Feder 1988. 'The Impact of Landowner Security: Theory and evidence from Thailand," The World Bank Economic Review 2(2): 187-204. Chand, Satish 1996. "Fiji within the CER," Pacific Economic Bulletin 11(1): 63-72. Chand, Satish 1999. 'The Public Sector and Development in the Pacific Islands," Pacific Economic Bulletin 14(1): 14-22. Chand, Satish and Ron Duncan 1997. "Resolving Property Issues as a Precondition for Growth: Access to Land in the Pacific Islands," Chapter 2 in P. Larmour (ed) The Governance of Common Property in the Pacific Region, Pacific Policy Paper 19, National Centre for Development Studies and Resource Management in Asia- Pacific, Research School of Pacific and Asian Studies, The Australian National University, Canberra, 33-46. 85 Corden, W.M., 1976. "Customs Union Theory and the Nonuniformity of Tariffs," Journal of International Economics, 6:99-106. Demekas, Dimitri G and Zenon G Kontolemis 1999. "Government Employment and. Labour Market Performance," IMF Working Paper 99/55. Due, J. F., and F. P. Greaney, 1992. "The Introduction of a Value Added Tax in Trinidad and Tobago," in Bird, R. M., and M. Casanegra de Jantscher, editors, "Improving Tax Administration in Developing Countries," IMF, Washington, D.C. Duncan, Ron and Xinpeng Xu, 2000. "Should Papua New Guinea Adopt a Stronger Exchange Rate Regime?" Pacific Economic Bulletin 15(2): 36-45. De Brouwer, Gordon, 2000. "Should Pacific Island Countries Adopt the Australian Dollar?" Pacific Economic Bulletin 15(2): 161-169. Duncan, Ron, Sandy Cuthbertson, and Malcolm Bosworth 1999. Pursuing Economic Reforms in the Pacific, Pacific Studies Series, Asian Development Bank, Manila. Duncan, Ron and Tony Lawson, 1997. "Cost Structures in Papua New Guinea," Discussion Paper No. 69, Institute of National Affairs, Port Moresby. Ebrill, Liam, Michael Keen, Jean-Paul Bodin and Victoria Summers 2001. The Modern VAT, International Monetary Fund, Washington D.C. Economic Insights Pty Ltd, 1999. "Review of the MSG Trade Agreement," Report prepared for the South Pacific Forum Secretariat. Edwards, Sebastian 1988. "Terms of Trade, Tariffs, and Labour Market Adjustment in Developing Countries," The World Bank Economic Review 2(2): 165-185. Eichengreen, Barry 2001. "Capital Account Liberalisation: What do Cross-Country Studies Tell Us," The World Bank Economic Review 15(3): 341-66. Filmer, R. J., and T. Lawson, 1999. "The Fiscal Implications for Forum Island Countries of Alternative Proposals for Regional Free Trade Areas," The Forum Secretariat, Suva, Fiji. Fiji Islands Ministry of Finance and Economic Development, 1997. "Investment Incentives Review," Suva, Fiji. Frankel, Jeffrey A; Ernesto Stein; and Shang-Jin Wei 1996. "Regional Trading Arrangements: Natural or Supernatural," American Economic Review 86(2): 52- 56. 86 Forsyth, David and Nii Plange 2001. "Social Impact Assessment of PICTA," Forum Secretariat, Suva. Forum Secretariat 2001. Pacific Island Countries Trade Agreement (PICTA), Suva. Forum Secretariat 2001. Pacific Agreement on Closer Economic Relations, Suva. Forteza, Alvaro and Martin Rama 2002. "Labour Market Rigidity and Success of Economic Reforms Across More than One Hundred Countries," The World Bank. Garrett, Geoffrey 2000. "Capital mobility, exchange rates and fiscal policy in the global economy," Review of International Political Economy 7(1): 153-70. Grynberg, Roman. 2000. The Pacific Island States and the WTO: Towards a Post-Seattle Agenda for the Small Vulnerable States. Paper presented at the Global Conference on the Development Agenda for Small States, London, 17-18 February 2000. Haltiwanger, John and Manisha Singh 1999. "Cross-country evidence on public sector retrenchment," The World Bank Economic Review 13(1): 23-66. Harberger, Arnold C 1988. "Policy making and Economic Policy in Small Developing Countries," Chapter 11 in Dombusch and Helmers (ed) The Open Economy: tools for policymakers in developing countries, Oxford. Hughes, A.V., 1998. "A Different Kind of Voyage," Pacific Studies Series, Asian Development Bank, Manila. IMF, 1990. "Tonga: A Revenue System for the 1990s," International Monetary Fund, Washington, D.C. Kiribati Minister for Finance and Economic Management, 2000. Speech to the Comprehensive Pacific Program, Port Vila, 5 November. Knapman, Bruce and Cedric Saldanha 1999. Reforms in the Pacific: an assessment of the Asian Development Bank's Assistance for Reform Programs in the Pacific, Asian Development Bank, manila. Laird, Sam, 1999. "Regional Trade Agreements: Dangerous Liaisons? World Economy, 22:1179-2000. Levantis, Theo 1997. The labour market of Papua New Guinea: a study of its structure and policy implications, PhD thesis, The Australian National University. Matusz, Steven J and David Tarr 1999. "Adjusting to trade policy reform," Policy Research Working Paper No. 2142, World Bank. Washington D.C. 87 Milner, Chris 1998. "Modelling labour market adjustment to trade liberalisation in an industrialising economy," The Economic Journal 108 (March): 509-28. Mitra P., 1992. "The Coordinated Reform of Tariffs and Indirect Taxes," World Bank Research Observer, Vol. 7, No.2, July, pp. 195-218. Rama, Martin 1999. "Public sector downsizing: An introduction," The World Bank Economic Review, 13(1): 1-22. Rama, Martin 2001. "Globalisation and workers in developing countries," Draft dated 24 July 2001, The World Bank. Rama, Martin and Kinnon Scott 1999. "Labor earnings in one-company towns: Theory and evidence from Kazakhstan," The World Bank Economic Review, 13(1): 185- 209. Schiff, Maurice 2002. "Regional Integration and Development in Small States," Policy Research Working Paper No. 2797, World Bank. Washington D.C. Scollay, Robert, 1999. MFN (Non-Preferential) Liberalization by Forum Island Countries, Paper prepared for the Forum Trade Ministers Meeting, 1-2 June, 1999. Forum Secretariat, Suva. Scollay, Robert, 2001. Regional Trade Agreement and Developing Countries: The Case of the Pacific Islands' Proposed Free Trade Agreement. Policy Studies in International Trade and Commodity Series No. 10. UNCTAD, New York and Geneva. Scollay, Robert, 2001a. "New Developments in Trade and Economic Cooperation among Pacific Islands Forum Members," paper presented to Pacific Update 2001 held in Brisbane on 18 July 2001 by the National Centre for Development Studies, The Australian National University. Scollay, Robert, John Gilbert and Darryl Collins, 1998. Free Trade Options for the Forum Island Countries, Report prepared for the Forum Secretariat, Suva. Statistics Department, Government of Tonga, Annual Foreign Trade Report 2000. Thomas, Vinod and John Nash, 1991. Best Practices in Trade Policy, Oxford University Press, Oxford. Vanuatu Minister for Finance and Economic Management, 2001, "Speech to the CRP National Summit," Port Vila, 5 November. Wacziarg, Romain 2001. "Measuring the dynamic gains from trade," World Bank Economic Review 15(3): 393-429. 88 World Bank 1987. World Development Report 1987, Oxford. World Bank, 1991, "Lessons of Tax Reform," Washington, D.C. World Bank, 1997. "Kiribati Customs Tariff Review," East Asia and Pacific Region Country Department HI, May, Washington, D.C. World Bank, 1998. "Enhancing the Role of Government in Pacific Island Economies," The World Bank, Washington, DC. World Bank, 2000. "Trade Blocs: A Policy Research Report." Washington, D.C. World Bank, 2000a. "Cities, Seas, and Storms: Managing Change in Pacific Island Economies," Washington D.C. Yang, Hsiu-ling 2000. "Education, Married Women's Participation Rate, Fertility and Economic Growth," Journal of Economic Development 25(2): 101-18. 89 IMAGING - Report No.: 24417 EAP Type: ER