(A)PS 13?o PoLIcy RESEARCH WORKING PAPER 1390 Regionial Integration Some say the Baltics should look to greater trade and the Baltics integration with the East, but Sorsa argues that opportunities are better for Which Way? trade with the West. Pintta Sorsa The World Bank Interntonal Econorrics Departmer.t F Iternatonal Trade Division NO December 1994 POI.ucY RESEARCHI WORKING PArER 1390 Summary findings Somc propose that the Baltics seek deeper trade * Numerous nontariff barriers with the East, slow and integration with the East to maintain existing trade flows unreliablc payments, unstable currencies, and barter and because the Baltics have had little market access to arrangements increase transaction costs and impede the the West. creation of more trade. Sorsa argues against such integration, proposing - Preferential trading with Russia or Ukraine entails instead that the Baltics improve trade relations with the the risk of increasing external protection for the more West, where market access is likely to be less and less of liberal Baltics. This risk is magnified by the relatively a problem. slow adjustment of Russia and other former Soviet Union After assessing factor endowments, and using a gravity republics and the faster reform in the Baltics. model, Sorsa predicts that more than 90 percent of Baltic The recent free trade agreement among the Baltics trade will be with non-former Soviet Union countries. allows countries to maintain independent external trade Initial exports are likely to be labor- and resource- policies, without creating the many administrative intensive goods, because it is easier to adjust to Western problems of a union. Free trade agreements will not only standards with those goods. Eut in the long run, the improve market access but may help lock in reforms at Baltics will have a comnparative advantage in skill- lome, vhich may help attract foreign investment. With intensive manufactures, as their years of schooling are liberalized trade, competition from liberal Estonia may among the highest in the developing world. (Exports of help reduce protection levels in Latvia and Lithuania. labor- and resource-intensive products, especially from After initial adjustment, trade with the West will Estonia, have already increased. Estonia is the most promote faster, more sustainable growth. Allocation of advanced of the Baltics in its transition to a market resources based on world prices, and transfer of economy.) technology, will increase productivity growth. Trade Sorsa predicts the Baltics will eventually trade mosdy with the West will probably also lower environmental with Europe. costs. She says the Baltics are unlikely to benefit from deeper OECD protectionism is unlikely to become an trade integration with the East because: insunnountable obstacle to more Baltic exports to the - The lower adjustment costs and the benefits of West. Recent statements about Europe turning its back maintaining viable industries resulting from sustained on the reforming East seem exaggerated, at least for the trade flows with the East are likely to be outweighed by Baltics. Their position as the former Soviet Union the cost of lost opportunities in the West. member most discriminated against by Europe is * Temporary preferential arrangements entail high changing, as they rapidly climb the various pyramids of administrative costs and are rarely temporary. access to European trade. * Preferential trade could mean slower adjustment and powerful lobbies against change. This paper - a product of the International Tradc Division, International Economics Department - is part of a larger effort in the department to monitor and advise on the trade policies of transitional economies. Copies of the paper are available free from the World Bank, t81 8 H Street NW, Washington, DC 20433. Please contact Jennifer Ngaine, room R2-052, exrension 37947 (48 pages). December 1994. The Policy Research Working Paper Series disseninates the findings of work in progress to encourage thce exchange of ideas about development issues An objective of the series is to get rbe fidings out quickly, even ifthe presentations ar less than fully polished. The papers rarry the names of the authors and should be used and cited accordingr. The fndings, interpretations, and concdusions are the authors' own and should not be attributed to the World Bank. its Exective Board of Directors or any of its mrenber countrie Produced by the Policy Research Dissemination Center REGIONAL INTEGRATION AND THE BALTICS - WVVCH WAY? Piritta Sorsa World Rank *The views in the paper are those of the author and do not reflect those of the World Bank Group, its Board of Directors, its management, or any of its member countries. he author wishes to thank Messrs. J. Hansen, A. HiUman, B. Kamini, V. Nehru, L. Alan Winters and P. Low for he4pfid comments. TABLE OF CONTENTS Summary ). () 1. Introduction .......................... 1 2. Comparative Advantage .......................... 2 (i) The Short Run ................... 2 (ii) TheLongRun ................... 5 3. Integrationwith the East ...................................... 11 3.1 ntegrationwith theFSU ...................... 12 (i) Existing trade flows ................... 12 (ii) Complementaity in demand ........................... 16 (iii) Levelofbafriers .................................. 17 (iv) Economic structures and speed of reform ................... 19 (v) Administrativeisus ................................ 20 (vi) Political economy ................... 21 3.2 Intra-Baltic Integrtion ...................... 23 (i) Existing tradew ................... 23 (ii) Complementarity of demand and structura indicators ............ 23 (iii) Level of protmeion ................................. 24 (iv) Pace of reform ................................... 24 (v) Admistrtiveissues ................................ 25 (vi) Political economy .................................. 26 4. Trade and Integration with the West .............................. 27 (i) Complnit .................................. 27 (ii) Level of protection ................................. 27 (iii) Administrative and poical economy concerns ................ 28 5. Conditions of Market Access in Main Westen Markets for the Baltics .... ..... 29 (i) EFTA Markets ................................... 29 (ii) TheEuropeanUnion ................................ 36 6. Conclusions .......................................... 41 Bibliography .......................................... 47 () Summary Benefits from either deeper Eastem integration on either a temporary or permanent basis are doubtful for the Baltics. Deeper integration with the East has been proposed to help adjustment by maintaining existing trade flows, or because of difficulties in market access in the West. This paper will argue that the Baltics should rather continue to make efforts for closer integration with the West in their emerging trade relations, and that market access is unlikely to be a serious problem. Given the preferential structure of market conditions in Europe, free trade agreements are also a key to better market access. Geography and existing factor endowments predict that the Baltics would trade mostly with Europe and have a comparative advantage in skill intensive manufactures. According to a gravity model over 90% of the Baltics' trade would be with non-FSU countries. Initial exports are likely to be labor- and resource-intensive goods, because of easier adjustment to Western standards in these goods. In the longer run, the Baltics are likely to specialize in skill-intensive manufactures - their skil levels measured by years of schooling are among the highest in the developing world. Trade relations with the East are best pursued in the context of multilateral trade rules and pursuing unilateral liberalization at home. Benefits from deeper Eastern integration within a temporary or permanent preferential trading area are therefore doubtful. First, lower adjustnent costs, and maintenance of potentially viable industries that could results from the maintenance of existing trade flows with the East are likely to be outweighed by the cost of lost opportunities in the West. Trade policy is second best as social policy - safety nets are likely to target assistance in a more efficient way. Second, temporary preferential arrangements also carry high administrative costs and their temporariness can be doubtful. Third, preferential trade could also mean slower adjustment and create powerful lobbies against change. Fourth, trade with the East is currently beset with numerous non-tariff barriers, that impede trade beyond preferential access reducing potentidal for trade maintenance. Slow and unreliable payments arrangements, unstable currencies and barter arrangements increase transaction cost and impede trade creation. These underline the importance of tackling these issues first. Fifth, a preferential trading area with much larger partners such as Russia or Ukraine could exogenize protection for the Baltics in a negative direction and bring a risk of increasing external protection for the relative liberal Baltics. This risk is magnified by the present different speeds of adjustment between the faster reforming Baltics and slower Russia and other FSU republics. The recently agreed free trade agreement among the Baltics can help trade liberalization among them. A free trade agreement allows countries to maintain independent external trade policies, which avoids many of the administrative problems of a union. Trade diversion is reduced by the existing free trade agreements with Europe and continued unilateral liberalization at home. It could also help create intra-Baltic trade in production to Europe. Competition from the liberal Estonia can also reduce protection levels in Latvia and Lithuania. (ii} Trade with the West beyond the initial adjustment will promote faster and more sustainable growth. Allocation of resources according to world prices, and transfer of technology will increase productivity growth. Trade with the West is also likely to mean lower environmental costs. In Europe, given the pyramid of preferences, free trade agreements are iumportant for market access. Given the structural and economic characteristics of the Baltics al: Western Europe trade creation is likely to exceed trade diversion. The FTA also brings security to market access and helps to lock in reiLorms at home, which in turn may help attract foreign investments. OECD protectionism is unlikely to become an insurmountable obstacle to more exports to the West from the Baltics. The fears expressed in a number of statements and recent studies of Europe tmning its back to the reforming East seem exaggerated, at least for the Baltics. Their position as most discriminated FSU members is rapidly changing as they climb the various pyramids of market access in Europe. Access to most EFTA markets has substantially improved with the recent free trade agreements. The Baltics have been quick to take advantage of these opporunities. Exports of labor- and resource-intensive products have increased especially from Estonia. Estonia is also most advanced of the Baltics in transition to a market economy. In the European Union the Baltics are still among the most discriminated non-OECD countries, but better conditions are under negotiation. 1. Introduction Is regional integration a viable option for the Baltics in their trade relations with each other and third countries? The role of regional trading arrangements in transition or in the long run in the former centrally planned economies has been subject to differing views. A preferential trading area with the East' and among the Baltics during transition has been proposed to ease adjustrnent by maintaining existing trade flows for industries with positive value-added at world prices (Tarr-Michalopoulos (1992b), van Brabant (1993), ECE 1990, 1992). Permanent integration with the East has been proposed (Ethier (1992) and ECE (1992)) to create efficient trade flows among "natural trading" partners. Doubts about market access in the West has also been used as argument for deeper integration among the countries in transition (Messerlin 1993). In practice the Battics have ignored these arguments and have made more efforts to integrate with the West than with the East2. Despite higher initial adjustment costs, integration with the West is likely to bring substantial benefits from increased exports, investment, transfer of technology, and spill-overs from exposure to modem laws and ways of doing business. This paper will discuss three options for regional integration in the Baltics - with the East, among the Baltics and with the West. Although trade with the East is important, the benefits for the Baltics from deeper integration with the East even within a temporary preferential trading area are doubtful. Free trade among the Baltics can complement Western integration. This paper will argue that the Baltics should continue to make efforts for deeper integration with the West. The paper will also show that regional trade arrangements are important for better market access in Europe, and that so far market access has not been a major obstacle for trading with the West. 'The paper uses East for Former Soviet republics, and West for the rest of the world. Former CMEA countries (Bulgaria, Czechoslovakia, Hungary, Poland, and Rumania) are included in the West, as their trade with the Baltics during the Soviet period was minimal. 2 At the time of writing the Baltics have entered into free trade agreements with a number of EFTA countries and with each other and have agreed on a gradual move to free trade with the European Union (EU) as of 1995. General trade agreements exist with some FSU countries. 2 The paper will first discuss the likely determinai.ts of short and long-term trade flows and comparative advantage of the Baltics. Likely net benefits from trade creation and trade diversion with each option of regional integration are then analyzed on the basis of indicators and conditions developed by theory and past studies on regional integration. These indicators are existing and expected levels of protection, complementarity of demand, differences in economic structures and factor endowments. In addition, administrative and political economy concerns influence the costs and benefits from regional integration. 2. Comparative advantage How will the Baltics fit to the world trade scene? What would be the direction and strueture of their trade in view of their geographic location and factor endowments? The short run is likely to differ from the long run. i) The Short run Initially, trade patterns with the West are likely to i) reflect past patterns of production and trade, and ii) the ease of adjustrnent in different industries to world prices and Box 1. Trade distortions in central plmaing - timber conditions. Resource-based goods or simple A good exampleiof disoned tade pamerns from central pbnning labor-intensive goods are the most likely is the wood and paper sector in Estonia. Soviet planners assigned two paper mills to Estonia. Thesc used imported.fir and pine candidates for initial exports. Adjustment to from Russia as raw materials for the paper Estonia exporled to other FSU republics. Estonia had plenty of local wood - bich - world prices and standards in heavy or which was exported as logs. Since independence the factories suffer from the rut-off of timber supplies from Russia. which capital-intensive industries can be more have been declared sfrategic. Exisdng techmology is unsuitable for local birch, which is unformoate as birch provides z higher difficcult and bear the highest adjustment lual4 paper than p and fir. costs. Efficient production of many products may require new investments to match differences in local and world standards, and to replace outdated, wasteful technologies. In resource-based goods the switch between markets is likely to be easier as they tend to be more standardized, and quality can be less important in sales. Initially incentives for continued trade with the East are highest in machinery for differences in standards with the West and undervaluation in pre-independence trade. Highest 3 incentives for Eastern trade are in the most undervalued goods - energy and raw materials (World Bank 1992a). The opposite is true for consumer goods, food and agricultural products, which were overvalued. But gradually the incentives change as trade with the East will also change as relative prices adjust to clnnges in prices and exchange rates. The Baltics emerged from the Soviet era with highly distorted production and trade patterns. Trade within the Soviet bloc was based on state orders at pre-determined prices commonly ignoring comparative advantages or local resources (Box 1). At the end of the Soviet era in 1990 over 90% of the Baltic foreign trade was within the Soviet empire. The Baltics produced machinery, food and chemicals for export mainly from imported inputs and energy. Trade with the outside world was minimal. The goods exported were similar to those in the intra-republic trade - machinery and food were most important items (Table 1). Adjustment to world prices and trade with the West is likely to be easiest in Estonia, because of the relatively lower share of capital intensive Soviet-style industries there. Consequently, Latvia and Lithuania would have the most difficult adjustment. The emerging trade patterns in 1992/93 correspond to the above predictions, although the data are subject to large margins of error'. Estonia was the most successful in shifting the direction of its exports and had the most diversified product structure in its exports to the West. In 1992 Estonia and Latvia exported more than half of total to the West, while Lithuania marnged slightly above a fourth (see Table 6). Labor-intensive goods (textiles) in Estonia and resource-based goods (food, wood, metals) in all three Baltic countries dominated Western exports. Machinery exports remained important in trade with the East (Table 2) especially in Latvia and Lithuania. The better success of Estonia in turning to the West is likely 3At the end of the Soviet period Latvia and Lithuania had the highest shares (about 30 %) of heavy industry in total production compared to Estonia (17% ). Food and light industries (mostly textiles) were each about a quarter of total industry in all three B21tic states (World Bank 1992a). 4Hyper-inflation, substantial changes in exchange rate and deficient reporting from lack of border controls makes the data for 1992 and 1993 subject to many valuation problems, and therefore should be taken as indicative only. 4 Table 1: Main Exports of the Baltics in 1990 Intra-Republic Extra-Republic Product % Product % Estonia Machinery 29 Food 31 Chemicals 19 Machinery 22 Light Industry 16 Light Industry 15 Food 15 Metals 13 Machinery 40 Machinery 53 Latvia Food 15 Food 16 Chemicals 14 Wood 7 Other (Communic.) 9 Light Jndustry 6 Machinery 44 Oil 41 Lithuania Oil 12 Machinery 37 Light Industry 10 Food 8 Food 8 Wood 5 Source: World Bank (1992a). Table 2: Structure of Baltic exports in 1992 or 1993(%) Estonia*) Latvia***) Lithuania**) Products West -ast West East West East Food 15 18 9 20 34 15 Wood 14 6 18 1 2 2 Textiles 24 11 9 16 11 21 Metals 16 2 15 3 7 3 Machinery 7 20 3 34 9 31 Other 24 43 46 26 37 28 Total 100 100 100 100 100 100 Memo: Value Mio.US$ 240 212 518 619 225 580 *) HS Classification, last 6 months of 1992. **) SITC Classifiction. ***) HS Classification. Source: Estonian Central Bank. COMTRADE. State Committee for Statistics of Latvia (1994). 5 to have been influenced by its more rapid progress in stabilization, more liberal trade regime and policy reform in general (PlanEcon 1993). A large share of Latvia's exports to the West is likely to be re-exports of petroleum and metals from other parts of the fonner Soviet Union (!SU). For example, Latvia has no petroleum production capacity. In Lithuania, stabilization has taken longer to achieve. ii) The Long-run. In the longer run, once property rights have been established, infrastructure improved, and adjustments to world prices made the patterns of trade are likely to change to reflect more closely their comparative advantages (Box 2). Deeper integration with the East at distorted Box 2. Determinants of Comparative Advantage prices could slow this process, while Tbe determinants of the patterns of international integration with the West is likely to speed it trade can be differences among counies such as in up.' Geography and relative factor ffactor endowments- or differences in technology; specialization in differentiated. products from endowments can give some insights on likely economies of scale and imperfect.competition;. or :geographic proximity. Factor endowments tend to developments in the future. Historical trade -:be-more.important.in tradce in more homogencous, or resource-based -products. Among countries at- patterns are unlikely to be repeated in the different levels of incomteldevelopment differences future, and are only presented as reference in technology can be t in explaining trade pattems. .Trade in differentiated products tend to. be (Box 3). more common among similar, higher income countries. Geography as-determinant of trade flows has no established theoretical basis, except perhaps benefits from proximity such as lower transport and Gravity models. Gravity model communication costs. The short-comings of gravity simulations indicate that the West should models with,existing theory are well-documented (for a survey see Baldwin 1993). Nevertheless, they tend to fit,weU actual trade patterns and can provide a useful indication of likely trade pattems in the - absence of reliable trade data. They. explain trade - . patterns with distance, size and level of GDP. 5This is not to say that the distorted FSU market is not worth exploiting. This 'niche" market is likely to offer much trading potential during the adjustment period. But the choice of deeper Westem or Eastern integration will effect the speed of this adjustment. 6 absorb over 90% of Baltics exports (Table 3),h The European Union (EU) (dominated by Germany) would absorb nearly half of the Baltics' total exports. Next in line is European Free Trade Association (EFTA) with about a fourth of total trade, with Finland and Sweden on top of the list. EFTA is most important for Estonia (26%), although its importance to Latvia (20%) and Lithuania (16%) is not negligible. The EE6 (Bulgaria, Hungary, Czech and Slovak republics, Poland, Romania) countries share was small (2%). This leaves minimal shares for the Eastern partners. The share of the other Baltic states is less than one percent, and that of the other republics of the FSU was 6% despite their large physical size. The low share is explained by the low income levels among the Baltics and their Eastern neighbors. The data excludes developing countries, but distance and low incomes are likely to keep their share modest in total. Compared to projected data, actual trade in 1992 with the EU would be well below its potential, while Russia and other FSU were well above the predictions especially in Lithuania. Differences in endowments. Relative factor endowments can give some indication on the types of products the Baltics are likely to be competitive in and who might be their main competitors. Most trade of the Baltics as small, modest-income countries is likely to be inter- industry trade, in which relative factor endowments are important. The Baltics are very close to the world averages in their relative endowments of all three main factors of production - land (N), capital (C) and labor (L) (Figure 1). Adapted from Anderson (1991) Figure 1 presents relative endowments of land (proxied as population per unit of arable land; (LIN) increases from N to L in figure 1) and of capital (proxied as income per 6The simulations used the Wang-Winters (1991) coefficients, 1992 World Bardc data on GDP and populations(Pop) and direct distance (Dist) between capital cities (x and i). The equation is Xix=-12.5 + 0.38(GDP/Pop)x + O.79*GDPx + 0.22(GDP/Pop)i + 0.8GDPi - 0.75*Dist.xi in log terms. Box 3. Inter-var period. Most imports and exports dozing die intcr-war period were with Ibe morc indusrializd Gemany and Unitcd Kingdom. In late thirties over 90% of L2rvia's and over 80% of EstniWs and Lithuania's exporis wet to Wstern Europe;.Same applied for imports. The agrrian Baltic staues devclope to successfulaguicultlal exportcrs to the industlIzcd Western Emurp. Despite the cxtrn;al constints to tradc. their trade pattem reflected iner-indusmy spcrializaton along die lines of casic compa_tvc advantge. The Balties exchanged food and. natura resources such as wood against manufacures such as coviuier goos-and. machinery with Western Europe. Food (eggs and butter) was die main export item in Lithunia and Estoai. while-Latvia exported mainly wood.. Geograpbic proximity: had a moderatc impact on trade flows during tie inter-war period. Trade among the Badcs.or with tie neighboring Russia-or Nordic countries was-small and eve'ndeclind during the period. Estonia was Jmst integratd with tbe Nonlic countries with the highest shar of trde with them. Potendalfor trae among these nstly agrarin ecouionijes (apart fronm Swen) with modest incomc levels and producing smilar goods as: tlh Balicswas small. Communimn cut.t swith -their traditioal trauldi paimcr- .and nmed the FSU ountrics to vcry dcosid-economies. Table: -W . Bac - Export(pr- .. 2 ...19. .. .. Estonia' . Ll a LEstonl la Lithuania Countries :Noidc C 14:;) . 30 . - - 1........ :0 ) - -): - -aIts0 8-:(5) 3(-.41)- i - :) Russia. (1) '7 :'- ::~~~~~~~~~~~~~. ;.., . . . . . . Product -~ ~ ~ O . - - ,-:.: -s AgriuimtI- 2:8-26) 12 ) 38) i 7 (3) - (- - (:). Wood -.19 (-) 4 .-).. . , 3 i 14 - 4 :4 11 - Textiles 33 (12) 25-2) 5 (32) 16(19) 01 .D- 12(20)- Mahinr -) - ) -. -(46) t-; 2) ; -(22) Sourc: Leaguc of Nations. Figure 1: Relative Resource Endowments a Lan\ L.a moacco Gr aap EflA SI La,DS~~~~~~~ -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Labor 0.1 0.5 1.0 3.Z lO 32 CauftaJ. Souzce: world Bank (1992), FAD (L99Z). UrODP (1993) 9 Table 3: Direction of Trade in a Gravity Model ( Projected Exports Actual 1992193 Exports Estonia Latvia Lithuania Estonia Latvia Lithuania EU 51 56 58 5 25 17 EFTA 26 20 16 30 9 5 -Finland 11 5 4 21 2 1 -Sweden 8 8 5 8 6 3 EEC 2 2 2 4 5 6 Other West 14 15 16 14 13 2 Balies I I I 3 6 7 Russia 4 4 4 21 30 28 Odter FSU 2 2 3 11 12 33 Source: COMTRADE, own esinmaxes. Estonian Misny of Tradc, Smta Cominiune of S otf Latyi (1994). capita; (CIL) increases from L to C in figure 1). Both are presented as ratios of world averages in log scales.' Compared to other developing countries, the Baltics would be relatively well endowed with natral resources and labor, but have much less capital than the resource poor, industrializing East-Asian countries (such as Korea, Hong Kong, Singapore, Taiwan, China) for example. Within Europe the Baltics' present factor endowments put them close to other Eastrn Europeans, especially the Visegrad (Czech and Slovak republics, Hungary and Poland) group. Russia and other FSU countries are relatively more abundant in land resources. The factor-endowment triangle does not differentiate by type of labor, which is likely to be an important determinant of trade flows in the Baltics. Comparisons of relative skill and capital endowments can give a better picture of the Baltics' relative endowments especially in services and manufactures compared to their main competitors. Figure 2 presents relative endowments of skills (measured by years of schooling) and per capita income for the Baltics, a number of developing countries likely to compete with them in world markets, and industrial 'It is difficult to find suitable pmxies for the factor endowments, and therefore the figure should be read as indicative only. The land proxy, arable land, is likely to overestimate agricultural resources at the cost of other resources such as minerals, oil, forests etc. Also the capital proxy (GDP) may overestimate the capital base in the FSU countries as much of this is likely to be worthless at international prices. Figure 2: Skills and Income Per Capita 13 - us 12 3 Ef ta 11- 0 10 - Oxl Japan Battics o 9 _ aaKorea OEE Russia a " 8 ~- E EE5 "U7 - 0 Greece OHong Kong 2 ~~~~Ukraine 6 - a 1 Portugal MaIa.sia 6 - a China 4 Thaitand 1 Singapore rNorth Africa (4) .3 - nw a- 0 4 8 12 16 20 24 28 (Tliousands) Per capita Income In 1991 In US$ source: World bank (1992), UNIP (1993) 11 countries. Despite possible differences in quality of schooling, years of schooling should be a good proxy of average skills of the labor force. The Baltics have one of the highest skill levels in the developing world (Figure 2). Their education levels are close to those of many industrialized countries - higher than Southern Europe, for example. High skill levels relative to per capita income suggest that the Baltic comparative strengths lie in skill-intensive manufactures and services. Their closest competitors would be Eastern European countries - especially the Visegrad group because of similarity of endowments. Beyond the initial adjustnent period Southern Europe or Northern Africa - other main beneficiaries of European Union (EU) preferences in market access - should not be in direct competition with the Baltics. Skill levels in North Africa are very low and they are likely to keep specializing in simple labor-intensive manufactures such as clothing much longer than the Baltics. Over time as capital accumulates the Baltics may start competing with the industrialing East-Asian counties. Among the latter only Korea has comparable skill levels to those of the Baltics. But the East Asians are a fast moving target. Russia, despite its high skill levels, is likely to specialize in more resource intensive goods, because of larger relative endowments of resources. If other resources such as minerals, oil etc. are included the relative advantage in resource-intensive goods is likely to increase fiurther. Russia could also become a competitor in similar skill-intensive industries. Direction and structure of trade is also influenced by the nature of trading agreements countries undertake. The following chapter is an overview of some of the options the Baltics face. 3. Integration with the East The benefits from regional integration have traditionally been measured by the net of trade creation over trade diversion. Past studies (for a summary see Nogues-Quintanilla 1993) have identified conditions or indicators that increase the likelihood of welfare increasing integration. These can be i) modest level of protection with third counties (reduces trade diversion), ii) low level of non-tariff barriers between partners (increases potential for trade 12 creation), iii) large existing trade levels (reduces potential for trade diversion), iv) complementarity in demand (increases trade creation), and v) differences in economic structure based on differences in factor enc.owments or in income levels (increase potential for trade creation). The extent to which a preferential trading arrangement leaves scope for unilateral liberalization can also be important. Administrative costs and the political economy of protection may influence the choice between customs unions and free trade areas. In economies in transition the case for preferential trading arrangements has distinguished between short and long terms issues (Tarr-Michalopoulos 1992a). The process of transition has its own arguments for regional arrangements. Preferential trade may help ease adjustment costs during transition and the survival of positive value-added industries. Benefits from temporary preferential arrangements during tranmition are related to shares of "East-dependent" industry (machinery) in pre-independence exports, likely terms-of-wrade losses from price adjustments, or speeds of adjustment in the partner countries. Below the economic, administrative, and political economy arguments of deeper integration with the East and among the Balkics are discussed in more detail. Past trade patterns are too distorted to serve as basis for analysis. Trade policies in the Baltics and its trading partners are also changing rapidly making assessment difficult. Recent economic and trade data in the East is patchy, but it is used to the extent possible. 3.1 Integration with the FSU i) Eisfing trade flows During central planning the Baltics were heavily dependent on trade with the FSU. This gives some weight to the trade maintenance argument. Lithuania and to some extent to Latvia with the highest shares of machinery in exports (Table 4) would have largest incentives to pursue preferential twade with the East. No information is available on the share of potentially viable industries in this trade. 13 As long as raw material or other input prices in the other FSU republics remain below world prices (including transaction costs), and infrastructure for trading with other countries is deficient, the Baltics may have an incentive to imnport from the East. A case in point is energy. For example, the switching of oil imports to non-FSU sources in Lithuania was first made difficult for the lack of necessary port infrastructure. The share of energy in total inter-republic imports was highest in Lithuania (35%) compared to Latvia (24%) and Estonia (17%) (see table 4). Dependence on Eastern raw-materials may also arise on technical grounds as standards with the West may be different. Estimates of initial terms of trade gains (losses, table 5) for the Baltics for maintaining trade at distorted prices show that largest "gains" would be in Lithuania Table 4: Structure of Baltic Trade with the FSU (1990) % Estonia Latvia Lithuania Imports Energy 17 24 35 Capital goods 39 35 33 Intermediate goods 31 31 23 Consumer goods - 8 7 Food 4 2 2 Exports Energy 9 4 20 Capital goods 28 37 43 Intermediate goods 27 31 20 Consumer goods 20 20 7 Food 16 8 10 Source: World Bank (1992a). Includes intra-Baltic trade. 14 most likely due to the large share of oil in its imports (35%)Y If the oil price estimate is too low, the gains are even larger. On the other hand, if revenues from the services rendered by the Baltics in the transit trade are included the "gains" from trade maintenance are reduced. Many of these potential "gains" from maintaining existing trade flows have already been reduced by actual policy developments. As terms of trade adjust, the incentive to trade with the East is reduced. Russia, for example, has started to increase prices of many raw materials close to world prices for many export products such as petroleum. On the import side the need for a Table 5: Estimates of terms of trade losses from price adjustments in intra-republic trade (1990) Trade Balance in Trade Balance in Terms of Trade foreign prices domestic prices loss Country (1) (2) (211)-i (exports/imports) (exports/imports) Estonia 0.92 0.62 -32% Latvia 1.07 0.81 -24% Lithuania 0.89 0.58 -35% Source: Calculated from data in World Bank (1992a). preferential trading area for trade maintenan is questionable, because the Baltics have low duties for many inputs, reducing the need for and the importance of preferences. At the extreme, Estonia has zero duties for most goods. Also the existence of other trade interventions reduces the potential benefits from preferential duties. The price incentive for Eastern trade is reduced by the existence of export taxes in Russia and other FSU countries for many raw materials, which raise their prices closer to world market levels. The status of these taxes in a preferential sThe estimates compare FSU-Baltic prices to estimates of world prices in 1990 given 1990 trade volumes. But the world price estimates used in the table are likely to be subject to large margins of error. The estimated world price for oil, for example in 125 % the domestic price. In other sources (Tarr-Michalopoulos 1992b) domestic oil prices in the FSU have been quoted as being a tenth of world prices. The data also excludes services, which for the Baltic as transit centers would improve the balances (Brown-Belkdndas 1993). 15 trading area is unclear. If the taxes reflect differences in prices of internationally traded goods and local goods, tax free trade with partners may also cause an outflow of these goods from countries with slower price adjustments (Latvia, Lithuania, Russia) to faster reforming countries (Estonia), or further to the West. This is already happening with food, oil and metals. If export taxes on raw materials are maintained, they already provide special protection to transforming industries with or without a preferential trading area. This is because the tax tends to lower the price of the raw material domestically below world levels. This would suggest that as a first step trade maintenance with the East should especially address existing non-tariff barriers . A preferential (temporary) trade agreement with the East to maintain potentially viable industries and to reduce adjustment costs during transition also has costs. First, a preferential trade arrangements within a common external tariff works like a (export) subsidy to these industries prolonging adjustment to world prices. Many of the existing industries are unlikely to be viable in the long run and their continued maintenance will be at a cost to the economy as the resources could be more efficiently used in other activities. In addition to higher returns, trade with the West at world prices has positive externalities such as transfer of technology. Trade with the East is also likely to have high environmental costs. Subsidizing trade with the East to maintain old machinery and factor use is likely to continue the environmentally wasteful practices in energy and other input use. It could be a subsidy to pollute as well. Trading at world prices and with modern technologies would reduce waste and pollution inherent in the old technologies.9 But again this is not to say that these industries should not exist or trade with the East. Only that maintenance of these industries within a preferential trading area with the east at distorted prices is likely to maintain the distortions longer and slow down incentives for adjustment to trade at world prices. The magnitude of the potential loss depends on the height of the preferential margin, and on the likelihood of the industries being viable in the future. 91he pollution cost of existing production (energy waste, use of dangerous chemicals or other hazards) has to be weighted against recycling costs of existing equipment and inputs, environmental costs of producing new inputs or machinery, their transport costs etc. 16 Without the preferences the selection by market forces would be faster. Low protection obviously would lower the efficiency costs as well. Second, the preferential arrangement can also create powerful lobbies against change and reform, which can undermine any temporary nature of the protection. Too many temporary protection arrangements have turned permanent (textiles in the West is a good example). Once established, the preferential arrangement may be difficult to dismantle as lobbies for continued protection would have been created. Third, the efficiency costs of a temporary union are increased, if its time-limited nature is not credible for investors. This may amount to substantial losses in potential output. The efficiency costs of maintenance of existing activities could be reduced by targeting the protection to potentially viable activities. This is not recommended as it runs great risks of failure in picking the winners. It would be difficult to indicate which sectors or companies to support either with trade preferences to the East or other forms of protection. Experience with targeted subsidies is mixed - the choice is best left to the market. It would be much better to proceed with reform and provide the enterprises a stable macro framework and appropriate business incentives. The trade maintenance argument for preferential trade with the East is also a choice on speed of adjustment. A preferential trading area would slow down adjustment to world prices. The above discussion suggests that the benefits from trade maintenance with preferences are unlikely to exceed the opport-anity costs and risks for permanent protection. This would depend on how the preferential area would function. A small preference within a free trade area is obviously less harmful than a highly protectionist common external duty for the presently relatively liberal Baltics. Obviously, if the option is unemployment, the costs of maintaining labor in the "old" inviable industries is lower. But using trade policy for social policy may not be most efficient and a better option would be to give direct assistance to affected workers. ii) Complementarity in demand Past trade patterns suggest some complementarity or excess demand between the more 17 industrialized Baltics and the more resource-intensive other republics of the FSU. Whether the complementarity will hold in the future remains to be seen. Past trade patterns are unlikely to reflect true comparative advantages of the former centrally planned economies. Existing factor endowments imply limited complementarity in the future. Similarity of skill levels between the Baltics and the larger FSU countries suggest some similarities (figure 2). On the other hand, the larger endowments of resources in general in tde other FSU can lead to different specializations between them and the Baltics in the future. Beyond trade in goods the Baltics had important links with other FSU republics in the services sector. Their geographic location is likely to maintain some complementarities vis-a-vis other FSU countries in the future. A union without border controls could make maintenance or development of these links easier. Much of the trade between the FSU and the West transitted through the Baltic ports with rail and road links to the Russian mainland. As Russian trade with the West expands its own port capacity is unlikely to suffice in the short run. This offers substantial potential for the Baltics as transit centers. The Baltics also have potential as a transit corridor for land transportation between the large St. Petersburg economic area, Finland and the rest of Europe. But these benefits are not sufficient to justify a union and can be resolved by simplifying border and other formalities. iii) Level of barriers Trade creation and diversion are influenced by the level of pre-existing trade barriers as well as barffers between the members in the regional trading arrangement. The level of emerging external protection among the Baltics and the rest of FSU is quite different. In Russia the average level of duties has been around 18% for preferential partners and the double for non- preferential ones. Export taxes and other trade restrictions are numerous (World Bank 1993). Recent problems with reform in Russia also suggest a more protectionist tr in policies. The Baltics have relatively moderate levels of protectionl'. IC Unweighted average duties in Latvia were 10% for MFN trade and 13 % for non-MFN trade (without specific duties), although dispersion of rates was high (range 0-35%). In Lithuania average duties were only 3%, although duties on many locally produced products were 10-25 % with zero rates on inputs. Both countries maintain some export restrictions, some NTBs. Estonia has zero duties on nearly all goods. 18 All Baltic countries have or have had discriminatory policies against trade with the East. In Latvia, for example, export taxes are higher on non-convertible currency exports and barter sales than in exports in convertible currencies. Latvia is also charging higher import duties (20% compared to 15%) on imports of some products from trading partners that do not have bilateral trade agreements with it. Until 1993 exports to non-convertible currencies in Latvia and Estonia were subject to the value-added tax (about 18%), which was zero rated for other exports. A Customs union between the Baltics and the rest of the FSU is likely to result in higher protection than at present in the Baltics on average. Any common external duty for the liberal Estonia would inply trade diversion. For the two other Baltics the amount of likely trade diversion would depend on how high the external duty would be set. Present differences in trade policies and bargaining power seem large enough to make negotiations on a common external duty difficult. The negotiating power of the three Baltic among the fifteen other republics of the FSU or with the much larger Russia is likely to be small. A free trade area with a more protectionist partner could also lead to trade diversion. Even a temporary preferential trading area with the East within a common external tariff is likely to result in higher levels of protection than the Baltic would maintain independently. Benefits from integration would also be reduced by other restrictions on trade. Export taxes and other restriction were mentioned above. The intra-trade between the Baltics and other FSU countries is also beset with many non-border barriers reducing potential benefits from preferential trade. Payments problems, currency instability, communications, transport problems hinder trade creation that affect maintenance of trade flows, and still result in large shares of barter (World Bank 1993). Trade preferences would be second best in solving these payments and barter problems. Although the situation has been improving recently, their solution is also likely to take some time, which limits potential for trade creation in the near future. Trade has also been affected by the macro-economic situation including monetary instability and the demand shock in the FSU countries. These reduce potential for trade maintenance or creation with or without preferences. 19 iv) Economic structures and speed of reform The differences between the Baltics and the rest of the FSU in the pace of reform and level of development may result in uneven gains and losses from integration. The Baltics have the highest per capita incomes in the former FSU, the fastest pace of reform and the most success in exporting to the West. The latter is often a good barometer of success in transition (Table 6). In many cases elsewhere in the world, maintenance of a union has required a successful system of transfers among members to compensate the less developed ones. The European Community, for example, has important transfer mechanisms in favor of its less developed but small southern partners. Depending how the FSU union would be constructed, the Baltics, that are relatively well-off members of the FSU, risk having to provide transfers to less developed republics. This can be politically very difficult and would be bad news for small Baltics supporting big Russia, Ukraine, etc. Table 6: Economic Indicators of the Baltics and some other FSU Countries GDP/capita Agriculture Exports to non-FSU as 1991 ,US$ in GDP 1991 (%) % of total, 1992 Estonia 3830 15 53 Lavia 3410 18 50 Lithuania 2710 30 28 Russia 3220 17 32 Belaras 3100 16 6 Ukraine 2500 24 10 Kazakhstan 2470 34 13 Source: World Ban (1992b), PlanEcon (1993). Different pace of reform among the members of a (temporary or pennanent) preferential trading area can even disrupt the process of reform. As for example Estonia has now largely liberalized most of its prices, competition from slower reforming Russia on subsidized goods could undenmine new investment at 'world prices", especially if export taxes in Russia were eliminated. The subsidies can be implicit in controlled prices, or in production of state 20 enterprises functioning at below marginal cost. The existence of export taxes in most other FSU republics suggests that substantial differences exist in the pace of liberalization of prices. The extent of the gains and losses depend on how divergent present policies are. The balancing of the gains and losses might require coordination of other policies such as competition policies, consumption taxes, explicit or implicit subsidies and liberalization of prices, or the setting of safeguards at the border. Negotiating on all these issues would consume precious time of the governments, which could be spent more productively elsewhere. v) Administrative issues. Preferential trading areas always imply some institutional arrangements, which can slightly differ according to the nature of the preferential arrangements. Both in free trade areas (origin rules) and in customs unions (sharing of revenue, customs administration, common external trade policy) these can imply important additional costs especially in countries with weak institutional capacities. In addition, as mentioned above both may require setting of safeguards, especially if the pace of reform differs substantially across countries, which carry their own risks for trade harassment. These additional costs apply to both temporary and pennrnent preferential trading areas. For the relatively small and liberal Baltics, the administrative costs of preferential trading arrangements are likely to be higher with a union than with a free trade area. Agreement on the common external duty can be time consuming and difficult. Revenme constraints in other members can increase pressure to raise the common external tariff. Distribution of revenue between the Baltics and the larger FSU partner could be difficult, and give rise to repeated disagreement among the members in the future. Members with better ports or other infrastructure (such as the Baltics) can become preferred ports of entry into the union at the expense of less developed regions and cause an uneven collection of revenue. If the Baltic states remain centers of transit trade, this problem is likely to be important. The scarce time and energy of the governments is likely to be better spent on other activities. 21 Free trade areas would have to face the issue of origin rules. These are required to prevent, for example, Japanese TVs, Hong Kong textiles or US bolts from entering Russia via Estonia free of duties. Preferential treatnent in a free trade agreement requires origin certificates and only applies to goods produced in the partner countries. This tends to give protection to producers of inputs in the partners to the agreements. in practice this can involve complex arrangements for companies for accounting of inputs and verification systems at the border and reduce the benefits from preferential border trade. The more different types of preferential agreements a country has with various partners the more complex the system may become. Origin rules can also be abused for protection depending on how they are implemented.In the absence of uniform rules, industry can easily negotiate tailor-made rules for their purposes and get very disguised protection. Strict origin rules can undermine preferential access to a partner market and reduce trade creation or trade diversion. By lobbying for tight origin rules, a protected industry in Russia can prevent a shift in demand to cheaper Estonian suppliers. This will reduce trade creation potential of a trade agreement. On the other hand, if a protected Russian producer does not comply with origin rules, Latvian buyers may be better off getting the same merchandise from third countries. This in turn reduces the trade diversion impact from the agreement. Cumulation of origin increases trade creating potential of free trade areas. vi) Political economy. The political economy concerns would work against a temporary or permanent preferential trading area between the Baltics and the East within a common external tariff. As the smallest and most liberal members of the FSU, the Baltics should maintain independence in external trade policies. A union with the more protectionist large partner carries a high risk of imposing the smaller ones to the protective structure of the large country. Preferential trade agreements can have both "positive and negative" effects on the political economy of protection. If protection is endogenously determined in a country, i.e. subject to lobbying, a preferential trading arrangement with several countries can result in lower protection by diluting the demand for it. This is because lobbying by one country becomes nore diffused, especially if the number of participants and industries is large. But a union with a more protectionist and large partner 22 may also exogenize protection in the negative direction, i.e. adjust it according to the lobbies of the larger countries. With a free trade agreement a country is free to set its external protection, but this affects only trade with non-partners. If a large share of trade is covered by the free trade agreement, the impact of increased border protection is small. Membership in an Eastern preferential area may also influence conditions of market access in the West. The West may be more cautious vis-a-vis preferential access for individual members of a customs union. Depending on the depth of integration within the potential Eastern union in which the member interested in preferences (a Baltic country) belongs, free access in the West to this one member could lead to free access to the whole union (FSU). The partners, the West and the Eastern union member (the Baltic country), would have to reach agreement on how to establish "member" (Baltic) origin from FSU origin, which can be difficult if the Eastern union has free internal movement of goods. Eastem partners may also be required to agree to the additional preferential trade partner in the West for one member. Depending on how deep the integration within the union should be, potential FTAs with the West would also have to be negotiated with the preferential union and not with the Baltic republics."' The presence of Russia or other FSU republics with slower reform carries the risk that the terms of these agreements could be much harder than for the Baltic alone. As small countries, the Baltics cause less fears for trade disruption. In summary, many economic factors point against deeper integration with the East. Trade diversion is likely to outweigh trade creation. The two regions are unlikely to be complementary in demand due to similar factor endowments, and deeper integration is likely to result in increased overall protection in the Baltics. Benefits from temporary preferences are likely to be outweighed by lost opportunities elsewhere in the economy, existing non-tff and non-border barriers in intra-trade, differences in the pace of reform, high administrative costs and the risk of creating lobbies against change. The exact balance depends on what type of integration is "In Latin America members of loose Customs unions such as the Andean pact have negotiated separate free trade agreements with other countries. But this requires that origin should be established. Presence of borders again reduces the benefits from union. 23 sought. The range from simple duty preferences to border-less free trade is wide. The case against a integration is strongest for Estonia with least dependence on the FSU, the most liberal policies, the most advanced reform process and a head start in trade with the West.12 Lithuania is the opposite case. But this does not mean that the Baltics should turn away from trade with the East - only that a priori there seems little benefit in subsidizing this trade. Security of market access and liberalization of partner policies can be pursued in the multilateral setting in the GATT. It also makes sense to tiot remove existing non-tariff barriers to intra-trade unilaterally before considering preferential trade arrangements. 3.2 Intra-Baltic Integration i) Existing trade flows The need for a temporary preferential trading area or a union among the Baltics to maintain existing trade flows is minor, because there was little genuine trade-dependence between the Baltic countries during the Soviet years. In early 1990s intra-trade accounted between 7-13 % of their exports. Most of the intra-Baltic trade was in energy products, and - even after the break-up of the Soviet Union and the coming into force of the Baltic free trade agreement - continued to take place duty-free. The Soviet planners attnbuted important roles for the Baltics in their centralized energy planning with large excess capacity at national levels in Lithuania and Estonia. The bulk of Estonia's exports to Latvia (75%) and imports from Lithuania (58%) in 1992 was energy. Lithuania's main export to Latvia was petroleum (40%) (Annex Table 1). Much of this trade is likely to continue along the lines of existing pipelines and distribution networks with or without a preferential trading area as long as Baltic prices remain competitive vis-a-vis outside suppliers and other distribution networks do not exist. ii) Complementarity of demand and structural indicators Trade creation in a Baltic preferential trading area is likely to be limited by the relatively similar nature of factor endowments in the three countries and the small size of their markets. A union with similar countries can be beneficial, if there is a large potential for intra-industry 12Sw PlanEcon (1993). 24 trade in differentiated products or with economies of scale. This is less likely to be the case with the labor-, skill- and resource-intensive products, in which the Baltic countries are likely to have some comparative advantage (Figures 1 and 2). Clothing, textiles, footwear or wood products such as furniture have limited scale advantages. Competition in many of these products is often based on flexible production structures and small series of rapidly changing products favoring small production units. This limnits the potential to create future comparative advantages with scale economies within a Baltic customs union. The small size of the potential internal Baltic market limits gains from trade creation within a Baltic union. The Baltics have 8 million people at relatively modest incomes. This makes thiem roughly the size of Sweden in population, but at 10% of its GDP. Trade in energy among the Baltics again is a special case. Economies of scale in this area can bring large benefits for the three small Baltic countries. Collaboration in developing common energy infrastructure, strategic self-sufficiency or rehabilitation or restructuring of existing facilities can be very beneficial for the Baltics. Most of this can be achieved without a preferential trading area. iii) Level of protection The differences in trade policies in the Baltics, although smaller than between them and the rest of FSU, are likely to make negotiations on a common external duty difficult. Any external duty for the liberal Estonia with practically zero duties would imply trade diversion. For Latvia and Lithuania free trade with Estonia could be trade creating. But it is doubtful that Estonia's Baltic partners would agree to a low and uniform common external tariff, which would be in the interest of Estonia. On the other hand, any compromise for Latvia and Lithuania would mean lowering of their duties and be trade creative. Transparency to the process of setting the duties could also improve in these two countries. A free trade area with independent external trade policies would be more workable for the three countries. Binding of policies in GATT could increase transparency of domestic policies. iv) Pace of reform Despite the relative similarities of the Baltic economies, the present differences in pace 25 of reform are sufficient to give rise to uneven distribution of gains and losses within a union. Among the Baltics, Estonia is the most developed of the three Baltic states with highest incomes and largest share of industry in output and fastest pace of reform (Table 7) making it a likely gainer within the Baltic union. Latvia is close to Estonia in its success of exporting to the West and share of industry in total output. Lithuania has had the slowest pace of reform and least success in turning to the West, although it is catching up. Estonia with its more open economy, convertible currency and stabilized economy might attract most new investments and take markets from the two other partners. A protected industry in Lithuania might lose to a more efficient Estonian industry within a moderately protected union. Transfer payments between Estonia and Lithuania seem politically unthinkable. The uneven distribution of gains could in time bring unnecessary friction among the members. A free trade area - already in place - should avoid many of these problems. v) Adminirative issues. Many of the administrative issues discussed with the FSU union apply to the Baltics as well. Agreement on common rules for the distribution of customs revenue within a Baltic union could be difficult. The Baltic countries have different revenue constraints. In Latvia and especially in Lithuania, pressures to raise revenue at the border are higher than in Estonia. Lithuania has stated an explicit target of collecting at least 10% of revenue from import duties. In the more liberal Estonia, at least in the past, revenue concerns have not influenced the level of duties, which have been kept almost to zero. The nature of origin rules in existing Westem FTAs favor an intra-Baltic free trade area. These allow the Baltic countries to cumulate origin among themselves in their exports to the Nordic partners. This may also become a rule with the EU, when the Europe Agreements are negotiated with the European Community. Cumulation of origin can contribute to trade creation by fostering intra-Baltic trade links in many intermediate- and input-producing industries. This is facilitated, if intra-Baltic trade is free of duties. If, for example, a Latvian manufactarer of textiles has to pay a 15% duty on cloth from Lithuania (the case without a Baltic free trade area), its incentive to cumulate inputs (purchase inputs in other Baltic countries) for duty free 26 sales to Sweden can be destroyed. This can be corrected with a duty-drawback scheme in Latvia, but probably at much higher administrative costs than a duty-free regime. With no duties, this problem does not arise for Estonia. vi) Political economy. Most political economy considerations favor a free trade area compared to customs union among the Baltics. A union among the Baltics might diffuse protectionist pressures in the less liberal partners, i.e in Latvia and Lithuania, by making local lobbying more difficult. For Estonia there is a big risk that the external protection would increase in a union. A free trade area among the Baltics in which one member is very liberal (Estonia) could reduce the power of the lobbies of the inefficient industries in Latvia and Lithuania as they lose markets to more efficient ones. This could contribute to overall reduction in protection there. In all three countries the risk of abusing emergency protection measures remams, in which case the endogenous nature of the protection would work against efficiency. The above suggests that, although a preferential trading area among the Baltics seems ill advised, the free trade area that the Baltics have now agreed upon can be a good option for them. The benefits are clearer for Latvia and Lithuania than for Estonia. An FTA can provide some trade creation among the Baltics as preferential suppliers to each other and can put them on equal footing with the Nordic countries with free trade agreements. Trade diversion will be limited as long as the Baltics maintain a low level of protection. A free trade area avoids many of the administrative problems of customs unions - revenue sharing and worries about distribution of gains and losses within the union. The unfair trade problems raised by the differences in the pace of reform could be solved by appropriate safeguard rules in the agreements. The Baltic countries can also cooperate in many other areas such as infastructure, designing of regulations and laws, etc. Baltic cooperation in building roads, communications links, and in developing transit trade between East and West can be very beneficial, although difficult to attain. Politically the Baltics can form a common front in negotiations with the West and East. 27 4. Trade and Integration with the West i) Compleruentarity Many structural and economic factors point to net benefits from deeper Western integration for the Baltic,. The large size and high incomes of the West European and the likely complementarity of demand between the Baltics and most West European countries offer large potentia! for trade creation. The differences in skill/capital endowments between the Baltics and most European countries in figure 2 are substantial. Deeper Western integration also offers substantial institutional and dynamic benefits for the Baltics. Integration with a m,ore developed partner will expose the less developed partner to modem laws and regulations, and can help transfer of technology. It may also help lock in reforms at home as a large share of imports has contractal limits to protection. Free trade agreements provide for larger security of market access than preferences given in the framework of the Generalized System of Preferences (GSP) or other unilateral schemes, which can be essential to attract foreign investment. The investors in turn can serve as a shield against emergency protection in the partner country. ii) Level of protection Efficiency costs from preferential trade with Europe are likely to be modest because of the relatively moderate average level of protection in most European countries (outside agriculture). As long as the Baltics continue with unilateral liberalization, the potential for trade diversion should be limited. This underlines the importance of unilateral liberalization before and after free trade agreements. Integration with the presence of trade barriers will, however, always imply some trade diversion from more efficient sources of supply, which has to be weighed against the benefits. Customs revenue will also be lower. These costs are likely to be lowest for Estonia with no duties. In Europe, integration in many cases is a pre-condition for better market access and against discrimination vis-a-vis main competitors. Market access is differentiated by various preferential agreements lowering both tariff and non-tariff barriers for exporters within a pyramid of preferential access (For more details see Pobl-Sorsa 1992). For example, in textiles the MFA is not applied to free trade partners, although some of these are subject to other but 28 less onerous restrictions. Market access in a number of sensitive goods is also more restricted. As the Baltics are initially likely to have some comparative advantages in sectors such as clothing, footwear and steel, preferential access can be important for export development. For example, Estonia with its zero-duty policies is nearly a text-book case of a free trader, but its export development depends largely on conditions of market access in the West. Without a free trade agreement with Finland, for example, it would face duties of 30-35% in many clothing items, which are one of its main exports. Box 4. The restrictive impact of origin rules in bilateral free trade agreements Origin rules can limit Baltic export potential to their Nordic partners. They favor resource-based products, but discrininate against specialization according to stages of production. To get duty free treatment for a garment in the Swedish market, for example, under the FTA the cloth would have to be of Baltic or Swedish origin, and the final product sold only in Sweden. This tends to give protection to producers of inputs in the two partner countries. If the Swedish fimal product is exported to other EFTA or EU countries the Swedish exporter would have to pay the duty on the part of the product that was imported from the Baltics. The agreements also require separate accounting of inputs according to origin or sales according to destination, which increases costs to producers. If duties are low, the preferential margin is eroded by administrative costs. EFTA estimates show that extra costs imposed by complying with the rules of origin can be up to 2 % of the value of exports (Herin 1986). iii) Administrative and political economy concerns Trade creation from deeper integration can be reduced by origin rules or by various forms of emergency protection in the partner countries (Box 4). Evidence on the restrictiveness of origin rules in the existing Baltic-EFTA agreements is mixed. For example, Finnish customs has not required origin certificates for the first nine months of the agreement between Finland and the Baltic states. Interviews with Estonian producers suggest that, especially in textiles, the local content requirement from the origin rules are very strict, and that they have difficulties in complying with them. These are areas in which future discussions of the agreements should pay careful attention. In the European Union the application of Anti-dumping duties against the Baltics will remain subject to more stringent rules as former communist countries, until they sign Europe types of agreements. Until mid 1994 only Estonia among the Baltics had faced one anti- dumping action in the EU. 29 The success of Western integration will largely depend on the willingness of the West to open their markets to the Baltics. In the following the importance of this potential constraint is discussed in more detail. 5. Conditions of Market Access in Main Western Markets for the Baltics. Is the West willing to welcorne the Baltics? Are the conditions offered such that deeper integration becomes possible? A recent paper by Kaminski-Yeats (1993) argues that the FSU countries in general have been among the least preferred suppliers in many OECD countries. The Baltics have done better than other FSU countries, but are still well below their most likely competitors - the reforming East European countries. The situation is changing rapidly, however. Access to main West European markets - EFTA and EU - is discussed in more detail below. Market access is also only one determinant of export performance. Even with most open markets inefficient producers are unlikely to succeed, which underlines the importance of domestic policies. Trade can also be influenced by other structural or natural factors such as natural resources or past structure of industry. i) EFTA Markets. Geographic proximity and high income levels make the Nordics important potential trading partneis for the Baltics and most EFTA countries have given them substantial concessions in market access. Bilateral free trade agreements have been in force since mid 1992 between the Baltics and Finland, Sweden and Norway. For Finland the agreement was a continuation of the duty free access granted previously to the republics of the Soviet Union. Agreement with Switzerland entered into force in April 1993. Austria, the remaining large EFTA member, has granted the Baltics GSP treatment. The bilateral agreements provide for duty free access for industrial products'3 subject to origin rules.'4 Agricultural products and fish (Sweden and Norway) are subject to separate agreements. "3The definition of industrial products in the agreements (HS chapters 25-97) is much wider than the traditional definition of manufactures in international trade statistics. The definition includes petroleum, many iron and steel products. "'Origin rules require that 1) a product is wholly obtained in the free trade area, 2) the tariff heading in the Harmonized System for the finished product must be different in the four digit level from the tariff heading for dte third country material used; 3) a specific percentage or processing requirement is met. The rules of origin as a general rule require substantial transformation to qualify as a local product. The large number of specific provisions on different products differentiates the rules. Textiles and clothing tend to have the tightest rules. 30 The EFTA Pyraniid of Market Access'5 It is typical for many European countries to differentiate market access with trade preferences related to either regional trade agreements or unilateral preferences. The Baltics are just below the European Union in the EFTA "pyramid of preferences". Other preferential agreements exist with Eastern Europe, Turkey and Israel, which entered into force in late 1992 and 1993. After the EU, the Baltic agreements are the most liberal in terms of market access, and contain practically no exceptions to duty free treatment or phasing in periods in industrial goods. Agreements with the Visegrad group, Romania, and Bulgaria, and Turkey and Israel have some initial restrictions on market access in sensitive products and some of the benefits are phased in over a number of years. The latter are modelled more closely to the agreements that these have with tfie EU. Another difference is that the Baltic agreements are bilateral, whereas the other EFTA agreements (EE6, Turkey, Israel) are EFTA- wide (except in agriculture). This makes the Baltic agreements constrained to one country, whereas the EFTA-wide agreements allow for better cumulation of products.6 The restrictiveness of the bilateral agreements will depend on how strictly origin rules are implemented. '5The EFTA pyramid of preferences is of a more recent vintage and simpler than the EU one (see below). Until recently, apart from the EU and EFTA free trade agreements, the only preferential agrement was the Generalized System of Preferences (GSP). Finland is an exception. During the Soviet period Finland maintained free trade agreements with most of the CMEA countries as well as with the Soviet Union. At present trade with other successor states of the Soviet union is on an MFN basis. Recent surge in free trade agreements was partly influenced by the need to adjust the EFTA countries trade policies to those of the EU as part of their EU accession procedure. But the EFTA countries also wanted to get similar treatment as the EU in markets such as the EE6 (Czechoslovakia, Hungary, Poland, Rumania, Bulgaria), Turkey and Israel: FTA came in force with Turkey on October 1, 1992; with Israel on January 1; 1993 with Czech and Slovak republics on July 1, 1992; Poland on April 1. 1993; with Rumania on May 1, 1993; and with Bulgaria and Hungary in 1993. "Cumulation means that use of materials originating in the producing country subject to origin rules are allowed in die calculation of origin. For example, Nordic bilateral FTAs allow for cumulation for materials originaing in the three Baltic countries and the Nordic partner country. Cloth from Latvia processed in Estonia will enter Sweden duty free, if it othierwise fulfills the transformation rules for origin. The EU's Europe agreements allow for cumulation between the partners. For example, materials from the EU, Hungary, Poland, Czechoslovalda are cumulable for Hungarian exports to the EU. Same applies to the EFTA wide FTAs. In the EU GSP schem cumulation of third country materials is not allowed. For example, EU materials do not qualify for the determination of origin in the Baltic. 31 Other developing countries benefit from national GSP schemes. The extent of GSP benefits varies among EFTA countries. According to Table 7 these are relatively limited. In 1991, for example, only 15% in Finland or about 1% in Norway and 21% in Sweden of imports from developing countries would have actually received GSP benefits. Level of protection in EFA countries. The structure of protection in the EFTA countries makes preferences valuable in sensitive sectors because of their higher level of protection. Average duties of manufacures in the EFTA countries are relatively low, ranging between 6-9%, which will be futher reduced by the Uruguay Round. The high level of duties in a number of sensitive, labor-intensive sectors in the Nordic EFTA countries implies a large preferential margin (Table 8) A number of sensitive sectors such as fish, textiles and clothing, and footwear are in addition subject to a range of non-tariff barriers increasing the protective impact of duties on third countries that remain outside the preferences. Non-tariff barriers, Table 7: EFTA GSP Benefits (1991) Importers Share of Imports from Share of Imports frk m Developing Countries Developing Countries that received Preferences Covered by GSP (%) % of GSP Covered % of Total Imports Austria 45 100 45 Switzerland 69 71 28 Finland 26 59 15 Norway 23 64 1 Sweden 32 66 21 Source: UNCTAD TD/B/SCP/3 especially in Sweden and Norway, affect resource and capital-intensive sectors as well (Table 8)I1. This increases the importance of preferences in market access especially in products that '7Data on non-tariff barriers should be read carefully. Latest data available is 1988 and the situation may have changed during the last five years. As most Nordic countries use MFA restictions on textiles and clothing, and the data shows very low levels of NTBs in these items, it may not portray very accurately the actual situation. 32 are likely to be of initial export interest to the Baltics. However, the protective impact of the external duty is reduced by the existing free trade agreements that the Nordics have with the less protected EU suppliers with a duty free access to the EFTA markets'8. For example, Finnish clothing producers compete not with imports with a 30% duty plus quota rents from the Multifibre Agreement (MFA), but with EU (and other EFTA) producers operating within their, often somewhat lower protective margins. Apart from fish, market access in agriculture is subject to high barriers (Annex Table 3) with very few preferences. The Nordic countries have one of the highest protection levels in agriculture in the world. Sweden and Norway have given few concessions to the Baltics in agriculture. The existing agreements lower border duties for very few categories of products, and the products remain subject to special agricultural levies. Whle better access could improve export potential in some products, preferential access to the highly protected EFTA markets in agriculture at well above world prices may have high efficiency costs. Free access to the high priced Nordic markets could draw resources to parts of agriculture in the Baltics, which at world prices would be uncompetitive. Any reduction in the Nordic protection would also imply additional adjustment costs to the Baltics. A similar, Dutch disease type, situation has arisen in some developing countries from preferences in sugar, banmas and beef under the Lome convention in the European Union. Preferential access to the EU at well above world prices has attracted resources into production of the preference goods in several developing countries. This becomes inviable once protection in the European union is reduced. Climate in the Baltics works against many comparative advantages in agriculture. The Uruguay Round and potential membership of the Nordic countries in the EU are likely make present high Ievels temporary. This does not mean that the Baltics shodd not negotiate better access conditions in agriculture, but that they should be aware of the potential costs involved. "gAssuming tat importers do not mark up prices to the EFTA protection levels. 33 Apart from better market access, the value of the FTAs for the Baltic lies in potential for foreign investnent. Nordic investors would also create a lobby against emergency protection in the North against the use of safeguard measures. A Nordic investor with interests in the Baltic is likely to lobby against the use of temporary protection in the Nordic countries against local lobbies for protection. But substantial amounts of investment are unlikely to materialize before property rights are more clearly secured and reforms have taken a stronger hold. The expansion of the services sector is also best promoted by foreign investment and know-how. To take full advantage of potential for transit trade, openness to both trade and investment is needed. Trade flows. Evidence from existing trade flows suggests that preferential suppliers are among the main trading partners of EFTA countries (Table 9). No doubt other factors such as geographic proximity and overall competitiveness are equally, if not more, important. EFTA countries traded most w.vith their EFTA and EU free trade partners, which account for dtree- fourths of total imports. The share of other preferential suppliers is small, which may partly reflect the relatively recent conclusion of most other free trade agreements. The three Baltics together accounted for less than 1% of imports in most EFTA countries in 1992. Israel and Turkey also have low shares, less than 1% of imports in most EFTA countries. The East Europeans were most important in the imports of Austria (5%) and Finland (2%) - no doubt a result of geographic proximity in the case of Austria, and past free trade agreements in the case of Finland. Actual trade performance suggests that the Nordic FTAs have helped market access from the Baltics, especially in sensitive products. The uneven performance of the different Baltic countries despite similar market access conditions also suggests that other policies also matter for trade performance. Most exports from the Baltics in 1992 were from Estonia to Finland and Sweden. Sweden was the main market for Latvia and Lithuania. Exports to Norway were small (Table 3). While some of the products exported are likely to be re-exports from other parts of the former Soviet Union (oil and metals), the Baltics managed to export other labor and 34 Table 8: Level of Tariff Protection In the Nordic Countries and the EU in Sensitive Sectors (Duties %) (NTBs*) Finland Norway Sweden EU Finlind Norway Sweden EU Lnbnr-intensive Tuxtlles(63) 19 2 10 3 0.8 1,0 Clothing(B4) 31 17 13 3 7.0 9.0 Footwcar[85) 14 7 12 6 3.0 67.8 94.8 Furniturm(82) 6 5 4 1 - - Resource-intensive FiHIOW3) 5 0.3 0.6 10 52.2 100.0 99,9 12.4 Wood(24+63) 2 1 1 1 - 53.9 63.2 10.5 Meals(28) 1 1 1 0 1.4 - - Cnpitral-Inteuslv Chemicals(S) 2 6 4 4 0.5 29.0 25.8 0.0 Machinery(7) S 5 4 3 8.0 23.5 Imn & steel(67+68) 4 3 4 1 5 S7.2 30.9 For Memo: Manufacturng 9 7 6 0.1 9.4 21.0 3.4 (5 to 9 less 67.68) Source: Smrt datbase. Dudes arm simple avenges for all imports, numbers in buckets are SITC, Rev I calegories of products. *-Nontariff barriers are percentage of imporls covered by all NTBs in 1988. resource-intensive products as wefl. Clotiing and textiles figured prominently in exports of all three Baltic countries to the Nordic EFTA members (Table 10) - up to 20-30% of their total exports. Another important export was wood, especially from Latvia. Despite the short life of the agreements so far the data shows that the agreements have enabled the Baltics to make a good start in exporting to the West. This is the more remarkable in view of the deep recession in the Nordic countries during 1992. EFTA consumers are also gainers as the trade displacement is likely to be at the cost of more protected, i.e. high cost, existing EU or EFTA producers. 35 Table 9: Main Preferential Sources of hniports in EFIA Countries (1992) % Share in Imports From: To: Finland Norway Sweden Austria Switzerland 1. EFTA 19 22 16 7 72 2. EU 46 49 55 68 7 3. Baltics 1 0 1 0 0 4. EE6 2 1 2 5 1 Israel 0 0 0 0 0 Turkey 0 0 0 0 0 5. All Developing 13 12 12 12 8 6. Russia 7 2 1 1 0 Source: UN Comtrade data base. Potential for future trade creation for the Baltics in EFTA markets can be substantial, which increases the value of preferential access from deeper integration. As skill levels in the Baltics are very high and wages relatively low, the Baltics could rapidly replace other preferential suppliers such as Portugal and Greece in labor-intensive goods in the Nordic EFTA markets in the intermediate run. Preferential access to the protected Nordic countries is likely to have benefitted Southern Europe at the cost of Lower cost developing country suppliers. For example, Portugal holds a substantial share in clothing imports of the Nordic countries, between 12-16% of total (Annex table 2). This is likely to change once the free trade agreements with Eastern Europe and Turkey become effective in 1993'9 and start taking hold. But differences in skill-endowments suggest that in the longer run both the Baltics and the Eastern Europeans '9EFor Finland this only applies to Turkey as the Eastern Europeans have had a free access in the past. Their share of the Finnish clothing imports is also relatively higher than those in Sweden and Norway. 36 would move to other products. Competition from the EE6 and Turkey with the Baltics is likely to be keener in Austria and Switzerland than in the Nordic countries because of geography. Even without FTAs, Eastern Europe has substantially increased exports to Austria (+87%) (Table 11). In other EFTA countries export success of the Eastern Europeans has been below average, except in Sweden. Turkey has increased exports everywhere. The potential membership of the Nordic countries in the EU as of 1995 may pose some difficulties for the maintenance of the Baltic free trade agreements. As members of the union they in principle are obliged to adopt the EU's external trade policies. The Baltics may lose if market access conditions of the EU are substantially different from the present EFTA agreements. The Baltics should use the time in interim to take maximum advantage of the present agreements to establish trading links with their Nordic partners. ii) The European Union The sheer economic size of the EU and its relative geographic proximity makes it the most attractive market for Baltic exports. However, at present the Baltics enjoy a much less privileged access to the EU than to the EFTA markets. As part of the FSU they were first subject to special trade restrictions aimed at communist countries, which where more restrictive than the simple most-favored-nation (MFN) treatment. In 1992 the situation changed and the Balties were granted regular MFN treatment and limited GSP benefits, which in 1993 were extended to textiles covered by the Multifibre Agreement (MFA). At the time of writing the Baltics have started negotiations with the EU for free trade agreements. The EU pyramid of Market Access. Access to the EU market is even more fragmented by preferences than that to the EFTA countries: i) the EU has free trade agreements with the Nordic countries on industrial goods and soon in services; ii) unilateral preferences with the 69 Asian and Pacific (ACP) countries under the Lome Convention include duty free access on most industial goods without non-tariff barriers and substantial preferences in agricultural goods; iii) association agreements with six Eastern European countries (Visegrad, Bulgaria, Romania) provide for gradual free trade in industrial goods and some benefits on agricultural products; iv) many of the twelve Mediterranean countries have association or free trade agreements with the 37 EU providing for duty free trade in most industrial goods with few non-tariff barriers and substantial preferences in agricultural goods; and v) GSP preferences exist with the ining developing countries and most countries in transition. Compared to their likely main compettors to date, the Baltics have been disadvantaged in access to EU markets. Level of Protection. The overall level of protection, especially in the sensitive sectors, is lower in the EU than in many of the EFTA countries (Table 8). While this reduces slightly the importance of preferences, the larger number of preferential suppliers can make them important for market access. Duties on sensitive sectors range between 3-6%, although these have to be increased by the impact of non-tariff barriers. Data on existing barriers on Baltic exports in not available. Table 12 gives some indication on the structure of barriers in 1990 for typical GSP (Poland) and1 non-GSP (Czechoslovakia) beneficiaries. Although the rates are biased by the countries' structre of exports, they give some indication on the differences in protection between different preferential categories of suppliers. The above table shows that a large part of existing Baltic exports should already enter duty-free under the GSP. Prior to its Europe Agreements, Polish duties as a GSP beneficiary, for example, were zero for many of these categories. This can be an overestimate as the SMART data used in the table assumes that Poland did not exceed the ceiling or quota limits in the GSP. This risk is reduced for the Baltics as small countries, because the ceilings and quotas in the EU tend to be insensitive to country size. For example, Russia, India and Estonia can export 6.6 million ECU worth of luggage or 15.4 million ECU worth of chairs and their parts to the Comununity duty free annually. However, many of the categories were subject to non- tariff barriers. Despite similar duty free treatnent as the GSP, the Europe agreements offer other important advantages. Free trade partners are likely to face lower non-tariff barriers in the EU, and get a 'fairer" anti-dumping review as mentioned above. Also coverage of duty free tratment in the sensitive categories in the Europe Agreements is likely to be larger than in GSP. In the long run, all trade in industrial goods within the "Europe agreements" should be duty free. The Europe Agreements also provide for wider cumulation of origin than GSP (see footnote 15), and larger security of markets access as GSP benefits are determined annually. 38 Table 10: Product Structure of Baltic Exports to the Nordic Countries (1992, % of total) Estonia Latvia Lithuania Fmland Food 6 0 0 Textiles 20 15 31 Wood 14 61 24 Metals 36 2 1 Oil 1 5 - Other 23 17 44 Value (mio. US$) 119.3 16.7 9.6 Sweden Food 3 1 4 Texdles 19 10 11 Wood 18 18 24 Metals 36 4 3 Oil 7 59 54 Other 17 8 4 Value (mio. US$) 97.3 90.5 53.9 Norway Food 1 0 0 Textiles 19 3 20 Wood 3 0 0 Metals 6 60 4 Oil 56 10 60 Other 15 27 16 Value (mio. US$) 8.1 10.5 2.9 Source: UN Comtrade data from importer data. The categories in SITC terms are Food=0+ 1+22+4, Textiles=26+65+84, Wood=24+62, Metals=28+68+67, Oil=3. 39 Table 11: Increase in EFTA Imports from New Free Trade Partners (percent over five years, 1992/1988 constant prices) Exports from: EE6 Israel Turkey World Imports by: Finland 5 -7 48 -1 Norway -9 -11 161 12 Sweden 25 20 10 9 Austria 87 37 59 48 Switzerland 0 43 18 16 For reference: EU 95 22 55 39 Source: COMTRADE Table 12: Baltic Exports to the EU in "Europe" Categories in 1992 and Level of Protection in the EU (1990) Level of Protection by product categorv Share in exports to EU (%) Duties (%) NTBs* "Europe" Product Estonia Latvia Lithuania GSP Non- GSP Non- Categories (Poland) GISP (Poland) GSP (Czech) (Czech) Al industrial 92 95 93 0.1 7 24 24 goods Share in ind.- goods: 16 0 3 6.2 4.4 15 40 12 2 1 3.3 3.1 13 100 One-year delayed 26 7 14 0 8.7 22 21 Two-year delayed 1 2 1 0.1 5.6 57 64 Quota/five year 7 4 4 0 10.7 89 88 ECSC 38 85 77 0.1 5.6 57 64 MFA 9 67 49 Immediate free trade - of which oil Source: Comtrade database. Duties and NTBs based on 1990 data, Baltic export data is from EU import data. * NTBs are percentage of imports covered by them. See footnote 18 for definitions of the categories. 40 Trade flows. The EU was a more important market to Latvia and Lithuania than for Estonia in 1992 (Table 3). This may reflect geography or the dominance of re-exports in trade with the EU. The lower market penetration of the Baltics in the EU in sensitive products than in the EFTA can reflect poorer market access conditions in the EU compared to EFTA. Other factors influencing this can be unfamiliarity of the EU market access conditions or lack of business contacts. The share of sensitive labor-intensive products in Baltic exports to the EU in 1992 was 10% or less compared to 20-30% to the Nordic EFTA countries (Table 13). Main exports to the EU were oil from Latvia and Lithuania, and metals from all three counties. A large part of this is likely to be re-exports from the other FSU republics. Projecting the impact of lower barriers in the EU on Baltic exports with existing export patterns could be misleading and of limited value. However, table 12 gives some indication of the likely initial impact of Europe-type agreementstm on the Baltic exports. It divides present Baltic exports intc the various categories according to conditions of market access in the Europe Agreements (Kaminski 1993). Well over 90% of their exports to the EU were industrial goods, which are covered by the agreements. According to the Polish example, duties for GSP beneficiaries in these categories would ahready be zero. Existing trade flows suggests that Europe agreements would be most benefLcial for Estonia. However, existing trade flows are unlikely to reflect future potential. Estonia had the lowest share (40%) in the inmediately free category compared to Latvia (85%) and Lithuania (77%). A large share of this in the latter two countries was oil. Estonia also has much higher shares of its exports in the more sensitive one-year (16%) and two-year delayed category (12%) than her Baltic neighbors. Between 12 and 34% of Baltic exports fell into the most restricted categories: quota/five year, coal and steel (ECSC) and textiles and clothing (MFA) categories. Steel and textiles have relatively small shares, which may have been influenced by the lack of preferential access in 1992. As non-GSP beneficiaries, the Baltics were subject to the high duties on these products. The example of Czechoslovakia (another non-GSP beneficiary) shows that duties on these three categories were on 3'Mhe Eumpe agreements negotiated with the Eastem European countries divide products to be liberalized into six groups with different speeds of liberalization for both tariffs and non-tariff bamers. i) one-year delayed group (duty-free access in the second year of the ETA); ii) two-year-delayed group (duty free access in the third year of the FlTA); ii) quotalfive-year delayed group (gradual increase of quotas and free trade in the sixth year of the FTA); iv) European Coal and Steel Community (ECSC) group (tariffs on steel eliminated by the end of the fourth year, tariffs on coal eliminated after one year, in Germany and Spain on the fifth year); v) MFA group (quotas gradually increased and eliminated after the fifth year); and vi) immediatel free trade residual. 41 average between 6-11% and a large share was subject to non-tariff barriers. With the inclusion of textiles to the Baltic GSP schemes in 1993, market access in these goods should have improved. Poland, for example paid no duties on these three categories. The above shows that market access conditions of the Baltics to the EU are being improved. Analysis of the Europe agreements compared to GSP shows that additional benefits in terms of lower duties would be modest, especially for present trade flows, but are important in terms of security of market access. The agreements can also greatly enhance foreign investments and thereby form a basis for future trade flows. The initial success of the Baltics in penetrating the Nordic compared to the EU markets in products with likely comparative advantages can partly be due to more difficult initial access conditions in the EU, although the situation is improving. It may also take more time to get familiar with the GSP as a preferential system. In their negotiations with the EU, the Baltics should use their size to their advantage. Market access is only one part of the equation. Progress with reforms at home is equally if not more important than market access for success in exporting. Estonia is most advanced in its reforms among the three Baltics in stabilizing its economy and initiating more substantial structural reforms, and this progress clearly shows in its initial export performance. All three Baltic republics faced roughly equal conditions of market access in the West, but Estonia is clearly the country that has not only succeeded in shifting most of its exports to the West, but also has the most diversified product structure of these exports. 6. Condusions Trade with the West is likely to promote faster and more sustainable growth than maintenance of existing trade flows with the East with preferences. Trade with the West is likely to promote investments and resource use closer to world prices, and is expected to improve productivity growth by transfer of appropriate technology. Trade with the West is also likely to mean lower environmental costs. In Europe - given the pyramid of preferences - closer integration is also important for market access, but also brings other additional benefits. Given the structural and economic characteristics of 42 Table 13: Share of Sensitive Goods in the Baltic Exports to the EU and the Nordics (1992) % Finland Norway Sweden EU Estonia Labor-intensive 26 20 22 11 Resource-intensive 53 56 50 40 Capital-intensive 12 6 15 20 Other 9 18 13 29 Latvia Labor-intensive 20 3 10 4 Resource-intensive 68 51 81 73 Capital-intensive 4 45 3 13 Other 8 1 6 10 Lithuania Labor-intensive 35 20 11 5 Resource-intensive 24 64 62 67 Capital-intensive 4 - 10 20 Other 37 16 17 8 Source: Comtrade. Labor-intensive=Textiles (65+84), footwear (85), fiurniture (82); resource-intensive=wood (24+63), fish (03), oil (3), metals (28); capital-intensive=chemicals (5), machinery (7), iron and steel (67+68). the Baltics and Western Europe, trade creation is likely to exceed trade diversion in a regional trade agreement. Integration also brings security to market access and helps to lock in reforns at home, which in turn may help attract foreign investments. Gravity models predict that the West would absorb most of the Baltic exports - over 90% of the Baltic trade would be with the non-FSU countries. Initial exports are likely to be labor-and resource-intensive goods, because of easier adjustment to Western standards in these goods. In the longer run, the Baltics are likely to specialize in skill-intensive manufactures - their skiU levels are 43 among the highest in the developing world. Main competitors are Eastern European reforming countries due to similar skill levels. Trading with the East can be important, but benefits from deeper integration with the East within a temporary or permanent preferential trading area are doubtful. Benefits from lower adjustment costs, and saving of potentially viable industries through the maintenance of existing trade flows with the East with temporary preferences are likely to be outweighed by the cost of lost opportunities in the West. The social cost of adjustment has to be balanced against the inefficiency cost of maintaining inviable industries and lost opportunities for growth. Trade policy is second best as a social policy - safety nets are likely to be better. Temporary arrangements also carry high administrative costs, their temporary nature can be doubtful as powerful lobbies are created against change. The underlying issue is largely related to the speed of transition and adjustment. Success of trade maintenance with preferences with the East can also be undermined by existing non-tariff-barriers. Slow and unreliable payments arrangements, unstable currencies and barter arrangements increase transaction costs and impede trade creation. Trade preferences would be second best in solving these problems. Under existing policies a preferential trading area with much larger partners such as Russia or Ukaine could also exogenize protection for the Baltics in a negative direction and bring a risk of increasing external protection for the relative liberal Baltics. This risk is magnified by the present different speeds of adjustment between the faster reforming Baltic and slower Russia and other FSU. The present free trade agreement among the Baltics can promote trade liberalization especially in Latvia and Lithuania. A free trade area maintains independence in extemal trade policy and avoids many of the administrative problems of a union. Trade diversion is reduced by the existing free trade agreements with Europe and maintenance of modest levels of protection. ft can also help create intra- Baltic trade in production to Europe. OECD protectionism is unlikely to become an insurmountable obstacle to the Baltics export development. The fears expressed is a number of statements and recent studies of Europe turning its 44 back to the reforming East seem exaggerated, at least for the Baltics. Their position as discriminated FSU members is rapidly changing as they climb the pyramid of market access in Europe. Access to most EFTA markets has substantially improved with the recent free trade agreements, The Baltics have been quick to take advantage of these opportunities. Exports of labor- and resource-intensive products have increased especially from Estonia. In the EU the Baltics initially were among the most discriminated non-OECD countries, but better conditions are under negotiation. In many products preferences are important for better market access. 45 Annex Table 1: Support to Temperate Agriculture* Wheat Milk Beef Eggs Pigmeat EC 77 200 107 0 8 Finland 328 483 271 149 183 Sweden 282 307 81 138 18 United States 70 138 40 8 6 Australia 17 42 9 14 4 New Zealand 10 3 3 118 0 * percentage of producer prices above world prices in 1990. Source: OECD Annex Table 2: Sources of Nordic Textile hnports in 1992 (%) Finland Sweden Norway EC Estonia 2.3 1 <1 0 Latvia 0.3 0.3 < 1 0 Lithuania 0.2 0.1 < 1 0 Portugal 16 13 12 5 Greece 3.7 3 3 3 Thailand 2.6 2 1 2 China 10.8 12 12 5 Hong Kong 7.5 9 8 7 E. Europe 7 3 1 4 Turkey 0.9 2 1 6 EC 51.60 50 59 47 EFTA 6 6 8 3 Source: Comtrade 46 Annex Table 3: Main Products in Intra-Baltic Trade in 1992 (%) Exports to: Imports from: Latvia Lithuania Latvia Lithuania Estonia a/ Electricity 75 Cotton 16 Cotton 10 Oil 58 Chemicals 8 Paper 10 Sugar 9 Paper 5 Fertilizer 2 Nuclear react. 7 Machinery 7 Cotton 4 Shareintctal 11% 2% 2% 3% trade Estonia Latvia Estonia Latvia Lithuania b/ Petroleum 40 Petroleum 40 Perfuimes 12 Textiles 17 Chemical - 10 Gas 15 Chemicals 12 Electricity 15 Textiles 9 Elec.Machinery 7 Dye 12 Transp.equip. 10 Share in total 2% 5% 1% 2% trade a HS classification (2 digit). b/ SITC classification (2 digit). Source: Estonian Ministry of Trade, Comtrade. Data for Latvia not available.