Implications of European experiences with regionalism for future economic integration in Asia

Transcription

1 1 Implications of European experiences with regionalism for future economic integration in Asia Richard E. Baldwin Graduate Institute of International Studies, Geneva First draft: November 2005; Final draft February INTRODUCTION European integration continues to fascinate and inspire nations around the world. By fostering economic integration, it brought peace and prosperity to a Continent that had been marred by almost continual welfare since the fall of the Roman Empire. It is natural therefore to search the European example for lessons. Lesson #1 is that European Union s visionary beginnings were critical to the group s success. A visionary beginning for East Asian regionalism would surely have been best, but the historical window-of-opportunity for visionary beginnings was missed. The concrete foundations of East Asian regionalism were poured without a plan. The East Asian trade building has two main wings; the FTAs between two major markets in the region the hubs, Japan and China and the smaller spoke economies in the region (the ASEANs). These FTAs are act as bilateral deals in the de jure sense for Japan and in the de facto sense for China. Laid over the two hub-and-spoke wings is an ineffectively implemented network of FTAs among the ASEANs (AFTA). AFTA mitigates some negative aspects of the region s hub-and-spoke bilateralism, but the disjointed nature of ASEAN and AFTA s unusual rules on preferences means that AFTA is not unitary. Its 45 bilateral trade flows are governed by 45 different tariff schedules, and only a small share of the intra-afta trade actually benefits from preferential tariffs (utilisation rates are under 10%). Despite its negligible commercial impact, AFTA is an important element of the building s foundation since it creates a forum for discussing and addressing issues that concern all the spoke economies.. In early 2006, Korea finds itself awkwardly perched between hub-status and spoke-status. It has come close to signing an FTA with Japan which would constitute a dominant element in East Asian regionalism, and it has come close to signing an agreement with ASEAN. Korea s role in the construction could be, but need not be, critical to the region s trade architecture. In short East Asian regionalism is marked by the Noodle Bowl Syndrome, i.e. a tangle of bilateral FTA, Two implications follow from this: Because East Asian regionalism developed at break-neck speed without any vision or plan, the highest priority for the region should now be management management of proliferating bilateralism and the Noodle Bowl Syndrome it implies. Further discussion of elegant architectural solutions that

2 2 might have been is intellectually stimulating, but what East Asia needs now is management, not vision. It needs what might be called a New East Asian Regional Management Effort, or NEARME. The complexity of the region s FTAs the Noodle Bowl Syndrome would pose a problem for any region in the world, but the problem in East Asia is much worse. Given the complex mesh of supplier relationships in the region, regional trade flows are the key to each East Asian nation s competitiveness; almost every trade flows in the region has the potential to affect the competitiveness of many major firms in the region. If something happened to disrupt intra-regional trade, Chinese firms would find it extremely hard to be competitive in the US market when it comes to timeliness, quality and price; the same can be said of Korean, Japanese and ASEAN firms. To paraphrase an old saying, if East Asians don t hang together on intra-regional trade, they may be hanged together in the world marketplace. East Asia, in other words, has a collective action problem. Because the need for management is pressing, the region would have to rely on existing institutional arrangements. It would take too long to set up a new organisation given the diverse political regimes and overlapping and intersecting interests in the region. While there are several contenders for the job, the ASEAN+3 body is probably the best suited. It is very easy to criticise ASEAN+3 for example it is unwieldy to have so many leaders in one place when they differ so greatly in terms of economic and trade importance but it has many positive sides. Chief among these is that it already exists, so all manner of delicate diplomatic issues can be avoided (just the question of who should participate could take years to resolve). Second, ASEAN+3 has a track-record of cooperation and thus has some credibility in the region. Third, it is not viewed as a creature of any single regional or non-regional power. Fourth, it has not to date been viewed as a threat to nations outside the region. Fifth, it encompasses the nations whose competitiveness in European and American markets depends upon intra-regional trade flows, i.e. the set of nations who have the greatest stake in ensuring that intra-regional trade is well managed. Sixth, it would be very easy to enlarge the group to include new members if and when the need arose. ASEAN+3 s institutional foundations, however, would need some firming up. A structure for the political direction of the grouping is already in place, but it would need a secretariat of its own with a high-quality staff of experts (legal, economic, customs and trade matters). Technically the new body would not be an international organisation (since making it one would take too long and involve too many conflicts), but the GATT operated as an international organisation from 1947 to 1994 without actually being an international organisation. Indeed, although it would be a very bad idea to call the new body an Asian WTO, given the fate of the Asian IMF idea, this is effectively what NEARME would be. The GATT operated on a consensus basis, engaged in soft law surveillance, enforcement and adjudication, and yet was the keystone to managing the rapid development of the world trade system in the post-war period. More to the point, the GATT s rules and peer-pressure approach is responsible for the fact that FTAs in Europe and elsewhere did not threaten the multilateral system. The GATT, in short, was the management effort that ensured the regionalism went smoothly in many other parts of the world. By focusing on regional issues and tailoring itself the East Asian sensitivities, NEARME could foster a level of management, i.e. discipline, in the region that is more solid than what the WTO offers under the Enabling Clause and more adapted to regional needs than the discipline it offers under Article 24. Plan of the paper The rest of the paper is organised in five sections. The next, Section 2, provides the background facts and a political-economy analytic framework for thinking about trade liberalisation in general and preferential trade liberalisation in particular. Section 3 presents the current state of play. Section 4 uses the political-economy analytic framework from Section 2 to organise thinking about the

3 3 causes of East Asian regionalism, with a discussion of the Asian version of the Domino Theory (Baldwin 1993, 1997) taking the central place. The Section also employs the framework to trace out the likely evolution of East Asian regionalism in the medium run. Section 5 discusses options for taking the region towards an integrated East Asia. This includes a more detailed discussion of the NEARME proposal, and a discussion of suggested management priorities. Drawing on the lessons from European integration, it also discusses longer-term institutional options for East Asian regionalism. The final section presents concluding remarks. An appendix presents a comparison of the pre-conditions for European regionalism in the early post-war period and East Asia today. 2. BACKGROUND: POLITICAL ECONOMY OF TRADE LIBERALISATION Regional integration involves preferential reductions in protection. The determinants of the scope of regionalism are therefore intrinsically tied to the political economy of protection. Although this logic can be extremely simple in simple cases, it can be quite involved in complex situation such as those in East Asia. It is therefore fruitful to review the basic political economy logic of protection before turning it application to Asian regionalism The quintessential logic of trade liberalisation From a political economy perspective, exports are good and imports are bad. This is why trade negotiators refer to tariff cutting as a concession when their own nation does it, but improved market access when their trade partners do it. Governments, in other words, are driven by mercantilist motives when it comes to trade policy. From an economic perspective, this view is either confused or just dead wrong 1, but the important fact is that this mistaken reasoning points governments in the right direction when it comes to trade talks. It leads them to conduct trade negotiations based on an exchange of market access, known as reciprocity in WTO jargon. Specifically, since exports are good and imports bad, if country A wants better access to country's B's markets, then country A is expected to 'pay' for this market access by opening its own market to B's exports. As we shall see, the combination of mercantilist motives and reciprocity unleashes a powerful political economy mechanism for trade liberalisation, what I dubbed the Juggernaut effect in my 1994 book, Towards and Integrate Europe. We start, however, by tackling the underlying problem of how governments would set tariffs if there were no WTO Political economy of unilateral tariff setting Imposing a tariff changes prices and thereby creates winners and losers within the tariff-setting nation. The political economy of tariff setting is all about the politics of these wins and losses. Suppose a nation s tariff is τ to start with and its government is considering a small decrease to τ minus ε. The tariff cut has three direct effects: it lowers the domestic price slightly, it raises imports slightly and it raises the import price slightly. As Figure 1 shows, the domestic price rises from P to P o, the import (or border) price rises from P -τ-ε to P o -τ and imports rise from M to M o. The lower domestic price raises consumer surplus by the area a+b+c+d, while it lowers producer surplus by area a ; the government s tariff revenue changes by -c+e. From a political perspective, the marginal political cost of this lower tariff is the loss to domestic producers, area a. The vast majority of protection in the world is put in place in order to boost the fortunes of domestic producers. In fact, even the name protection reveals its political intent. The 1 Imports expand the quantity of goods that are available for consumption or investment; exports reduce it

4 4 higher tariff harms one group of citizens (consumers) while helping another group (producers), so it might seem a bit strange to call tariffs protection. If governments cared a lot about consumers and nothing about producers, maybe tariffs would be call taxation. But since governments around the world tend to care more about producers than they do about consumers, tariffs are call protection. Figure 1: The winners and losers from a tariff increase. Price Price b d Po a c P P -τ-ε P o -τ S e τ ε D τ XS MD Change in imports Quantity M M o Quantity Olsen s asymmetry Why do governments typically care more about producers? The standard answer is known as Olsen s asymmetry. Governments who want to get and keep power typically adopt policies that help them do this and this usually means adopting policies that please politically powerful domestic groups. The precise mechanism by which the backing of power special interest groups helps the government get and keep power varies according to each nation s political architecture. But why is it usually the producers who form pro-tariff special interest groups? Why don t consumers band together to fight tariffs? As Mancur Olsen pointed in his famous book The Logic of Collective Action, import competing firms are much easier to organise politically. The firms that benefit from protection are relatively few in number, are often geographically clustered and each cares a huge amount about the tariff. Consumers, by contrast, are many in number, geographically disperse and care relatively little about the tariff on any given product. Indeed, most consumers in most nations have virtually no idea how much more they are paying for their goods due to protection after all there are many things in life that merit more attention than the nation s tariff schedule. This ignorance and lack of political organisation on the part of consumer dampens the extent to which tariff-setting politicians care about the loss that a higher tariff imposes on consumers. This is Olsen s asymmetry in a nutshell. Producers find it worthwhile to organise and push for higher protection; consumers don t, so governments pay asymmetric attention to the winners and losers from tariffs. Politically realistic objective function A parsimonious way of capturing Olsen s asymmetry is to assume that the government puts a higher weight on producer surplus than consumer surplus (and government revenue) when it considers the political pros and cons of raising the tariff.

5 5 It proves useful to divide the impact of a tariff liberalisation into two parts: the marginal political cost (MPC), which is the loss to producers, and the marginal political benefit (MPB) which is the gain in consumer surplus net of the change in the tariff revenue. Using Figure 1 we can plot how the MPB and MPC change as the tariff falls. As usual, the optimum in this case, the politically optimal tariff will be where the marginal benefit just balances the marginal cost, i.e. MPB equals MPC. Figure 2 shows a typical outcome. When the initial tariff level is very low, say τ 1, the marginal political cost from cutting it a bit, MPC, is higher than the gain from cutting it, MPB. Thus we know that τ 1 is not the politically optimal tariff. There are two reasons why the MPC curve is below the MPB curve at very low tariffs. The first is the classic optimal tariff argument that a nation can gain from imposing a small tariff since a small tariff forces a small terms of trade improvement while the distortionary impact is negligible. Second, the extra large weight on producer surplus in the Politically Realistic Objective Function reinforces the political gain from protection. Next suppose the tariff is very high, say τ 2, then marginal political benefit of cutting the tariff a bit outweighs the marginal political costs of doing so (the very high tariff starts to reduce the tariff revenue collected, i.e. the nation is on the wrong side of the Laffer curve and the Harberger triangles are very large). The politically optimal tariff is where the MPB and MPC cross, shown as τ in the diagram. Figure 2: The political equilibrium tariff in the PFS-lite model benefit/cost MPB (of a tariff cut) MPC (of a tariff cut) τ 1 τ τ 2 τ (tariff) If the import competing firms get stronger politically, the MPC of tariff cutting shifts up and the politically optimal tariff rises. If consumers organise or complain more loudly about the tariff, the MPB curve shifts up and the political optimal tariff falls The political economy of trade talks: the liberalisation paradox The discussion above serves to stress the fact that tariffs are not randomly chosen by governments. Tariffs are the outcome of a hard fought political process within each nation. But this then raises the liberalisation paradox : Why would governments ever find it politically optimal to remove tariffs that they previously found politically optimal to impose? The answer is that trade negotiations based on reciprocity change the array of political forces in each nation and thus change the politically optimal choice of every participating government. It is easiest to exposit the logic in the context of the post-war history of trade liberalisation. Wind back the clock to 1947 when the GATT became active. At that moment, tariffs started at levels that were politically optimal from a unilateral perspective (basically the tariffs in place at the

6 6 depth of the Great Depression). That is, tariff levels like prices in a competitive market balanced the marginal political cost and benefit of protection. At the politically optimum tariff, importcompeting firms and workers they employed would like the tariff to be higher, but the government is reluctant to raise tariffs further since this would impose too much harm on the general economic health of the nation. Starting from this situation, announcement of multilateral tariff-cutting talks based on the principle of reciprocity altered the array of political economy forces inside each and every nation participating in the talks on the basis of reciprocity. Reciprocity is the key. Reciprocity turns each nation s exporters from bystanders in the tariff debate to fervent opponents of protection within their own nation. Why? The principle of reciprocity means exporters win the prize of better access to foreign markets only if home tariffs are lowered. Thus lobbying against domestic tariffs becomes a way of lowering foreign tariffs. Importantly, this option was not available to domestic exporters before the reciprocal trade talks. To illustrate the importance of reciprocity, consider a multilateral trade negotiations (MTN) that was held without the principle of reciprocity. The chairman would simply gather all WTO members in a room, ask them to reflect for a few minutes on their tariffs and then to fill in a form listing the tariff cuts that they wanted to make. Since each nation s tariffs are already politically optimal from a unilateral perspective, all the forms would list zero tariff cuts. In the real world, however, the MTN works quite differently. The WTO members are allowed to bargain for tariff cuts. This exchange of market access has an entirely different outcome. In a sense, it allows each nation s exporters to lobby against foreign tariffs by lobbying against their own nation s tariffs. To consider more carefully the economics behind this political economy logic, turn to Figure 3. The usual problem facing a government is how much to cut tariffs, but the geometry is easier when we consider full elimination, so Figure 3 looks at the impact of complete tariff elimination. Figure 3: The economic impact of reciprocal tariff cutting. P 1 P a Home Market S b c d e τ f h P 1 -τ P 1 -τ MD D D* Q P World Market XS Imports P g Foreign Market C* Z* j S* P FT Q Tariff cutting has all the domestic effects discussed in Figure 1, but Figure 3 also shows the impact on foreign exporters. Elimination by the Home government of its tariff creates winners and losers in the Foreign nation. Foreign producers (i.e. exporters) gain since they see a higher price for their output and they get to export more. In other words, the change in Foreign producer surplus is the area f+g+h+j. The higher prices harm Foreign consumers by the area f, but since production exceeds consumption in the Foreign nation (that s why they are exporting the good), the Foreign nation as a whole gains from Home s tariff cutting. Moreover, presuming that Olsen s asymmetry is working in the Foreign nation as well, the gain to Foreign producers will be highly valued by the Foreign government.

7 7 But why is it important to consider the impact of Home s tariff cutting on Foreign, when Foreign firms and Foreign consumers cannot vote in Home? The reason is that the impact of a Home tariff cut on Foreign in Figure 3, tells us what Foreign tariff-cutting would do to in Home in the markets where Home is the exporter. To be concrete, suppose that Figure 3 shows the market for good 1, which is the good that Home imports from Foreign. And suppose that Home exports another good, say good 2, to Foreign. To keep things really simple, suppose that the market for good 2 is exactly like the market for good 1 with the roles of Home and Foreign swapped. This bit of mental gymnastics allows us to work out the impact of a reciprocal tariff cut. Under WTO rules the principle of reciprocity in MTNs in particular a nation that cuts its tariff can expect its trade partners to reciprocate. Reciprocity changes the political calculation facing the Home government. If the Home government eliminates its tariff on good 1, Home consumers gain area a+b+c+d but Home import competitors lose area a; to this we add the usual tariff revenue effect, -c-e. The new part is the gain to Home exporters that comes with the reciprocal elimination of the Foreign tariff on good 2. The Foreign tariff cut boosts Home exporters producer surplus by f+g+h+j, but lowers Home consumers welfare by f. Table 1 shows the political calculus: Table 1: Change in political objectives from removing the tariff. Without reciprocal MTN With reciprocal MTN good 1 good 2 good 1 good 2 Political cost (1+α M )(-a) n.a. (1+α M )(-a) (1+α X )(f+g+h+j) (weighted change in producer surplus from liberalisation) Political benefit (weighted a+b+d-e n.a. a+b+d e -f change in consumer surplus plus change in tariff revenue) Net -aα M + b+d-e n.a. -aα M + b+d-e (1+α X )(g+h+j)+ α X f Note: The α M and α X indicate the extra weight that the politically-motivated tariff-setters place on Home firms producer surplus. Without reciprocity, the government may or may not find it politically optimal to eliminate its tariff. Since most nations would impose some tariff if not constrained by their partners, we presume that the extra weight placed on firm s benefits is large enough so that the net -aα+b+d-e is negative; this means the Home government would not want to unilaterally eliminate tariffs. With reciprocity, the net benefit of eliminating Home is definitely larger since the gain from the reciprocal reduction in good-2 tariffs is unambiguously positive for Home, namely (1+α X )(g+h+j)+ α X f. What this tells us is that Home is much more likely to cut its tariff in the context of an MTN, and it is exactly reciprocity that alters the political calculus Juggernaut theory of reciprocal trade liberalisation 3 Since this rearrangement of political forces goes on inside every nation involved in the talks, a new political equilibrium emerges. Graphically, the announcement of an MTN lowers the marginal political cost of tariff-cutting (since now gains to Home exporters offset loses to Home importcompeting industries); this is a shift to the dashed MPC line in Figure 4. There is also some reduction in the marginal political benefit of tariff cutting since the terms of trade gain the Home 2 Note that if the alphas for imports and exports are both zero, the gain from full eliminator of tariffs is always positive; the basic point is that the terms of trade effects wash out so all that remains are the dead weight losses. 3 The Juggernaut theory was first posited informally by Baldwin (1994) and Baldwin and Baldwin (1996) and formalised by Baldwin and Robert-Nicoud (2004).

8 8 experiences on its exports creates some losers among Home consumers; the price of the export good rises and this lowers Home consumer surplus somewhat (MPB shifts down to the dashed curve). Normally, however, the drop in the MPB curve will be less than the drop in the MPC curve, so the new equilibrium tariff is lower. 4 The net outcome of the MTN is that tariffs get cut in all participating nations, although they need not go to zero. These tariff cuts are typically phased in over a 5 to 10 year period. Of course, this liberalisation alters the import-competing industries in all nations and the export industries in all nations. And this, in turn, alters the political economic landscape in a way that generates what could be called political economy momentum. The economic source of the political economy momentum In all participating nations, export sectors expand output and employment because the foreign tariffs cuts improve the exporters market access. The nation s import-competing sectors experience the opposite effect; they reduce production and employment as the home tariff cuts reduce the level of protection from foreign competition. In economic terms, the long-run supply responses in the export and import-competing sector are greater than the short-run responses. Why does this matter? Figure 4: Liberalisation with reciprocal MTNs. benefit/cost MPB (of a tariff cut) MPC (of a tariff cut) τ τ τ (tariff) Although many factors affect a sector s political influence in tariff-setting, size always matters. That is, other things equal, a sector with lots of workers, lots of capital and lots of profit will have more influence. Since the export sectors in all nations have expanded while all of the importcompeting sectors have shrunk the next GATT/WTO Round leads to a different outcome. A few years down the road, when another multilateral Round is launched, reciprocity again realigns the tariff-setting balance by turning exporters into anti-protectionists. But this time, the protariff camp is systematically weaker in every nation and the pro-liberalisation camp is systematically stronger in every nation. The result is that all participating governments again find it politically optimal to cut tariffs, but again not necessarily to zero. As these fresh tariff cuts are phased in the cycle is repeated. 4 The formal condition is that -aα+b+d-e +(1+α X )(g+h+j)+ α X f from Table 1 is positive.

9 9 This is the Juggernaut effect; once reciprocal trade liberalisation starts, it continues until all tariffs are gone. Any sector in any nation that is included in the reciprocal trade talks will eventually get liberalised. To study this more carefully, we note that the phased in liberalisation causes more changes than those shown in Figure 3. As Home lowers its tariff, some Home firms in the import-competing sector (what we called good 1) will exit and this shifts the Home supply curve to the left; simultaneously, the cuts encourage Foreign entry so the Foreign supply curve shifts to the right. Figure 5: Tariff cuts shift the supply curves (short- versus long-run supply elasticities). P Home Market S S P World Market P Foreign Market P 1 XS S* S* P 1 -τ P 1 -τ MD D D* Imports Q C* Z* Q P FT Now if we re-apply the calculus from Table 1 to the new situation, we will see that all the nations participating in the next round of MTNs will find it politically optimal to lower their tariffs even further. In a nutshell the Juggernaut effect describes the process by which reciprocal trade talks lead to tariff cuts globally and how these cuts themselves alter the political economy landscape in a way that foster further tariff-cutting. In particular, the global liberalisation strengthens the economic and political economic strength of the proponents of further tariff cutting (exporters) while simultaneously weakening the economic and political economic strength of the opponents of further tariff cutting (import competitors). Because this rearrangement of political costs and benefits goes on in each participating nations, the ultimate outcome is zero tariffs in all nations Reality check Oversimplifying to make a point, tariffs in the world can be summarised in three points: they are zero on industrial goods imported into rich nations, they are high on food imported into rich nations and they are high on industrial goods imported into poor nations (see

10 10 Figure 6 and Figure 7). How does the Juggernaut theory account for these facts? Until the 1986, GATT Rounds focused almost exclusively on manufactured goods, so it is no surprise that food was not liberalised. More specifically, the import competitors in the agricultural sectors in the EU, US and Japan were so strong that they managed to take food tariffs off the bargaining table. Consequently, no one inside the EU, US or Japan gained from lobbying against food tariffs. In this way, food tariffs were shielded from the relentless crushing force of the Juggernaut effect.

11 11 Figure 6: Tariff in rich nations. Chart 1 Simple average MFN tariff rates for the "Quad", by HS Section HS Section United States 2002 European Communities 2004 Japan 2004/05 Canada % 10% 20% 30% 40% 50% 60% 01 Live animals and prods 02 Vegetable products 03 Fats and oils 04 Prepared food, etc. 05 Mineral products 06 Chemicals and prods 07 Plastic and rubber 08 Hides and skins 09 Wood and articles 10 Pulp, paper, etc. 11 Textiles and articles 12 Footwear, headgear, etc. 13 Articles of stones 14 Precious stones, etc. 15 Base metals and prods 16 Machinery, etc. 17 Transport equipment 18 Precision instruments 19 Arms and ammunition 20 Miscellaneous manufs 21 Works of art, etc. Note: Source : Calculations include AVEs where available; where they are not available, the ad valorem part is used for alternate and compound rates. Excluding in-quota rates. WTO Secretariat calculations, based on information provided by Members.

12 12 Figure 7: Tariff in developing nations.developing nations have participated in GATT/WTO Chart 2 Simple average MFN tariff rates for selected developing countries, by HS section HS Section China 2002 India 2001/02 Brazil 2004 South Africa % 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 01 Live animals and prods 02 Vegetable products 03 Fats and oils 04 Prepared food, etc. 05 Mineral products 06 Chemicals and prods 07 Plastic and rubber 08 Hides and skins 09 Wood and articles 10 Pulp, paper, etc. 11 Textiles and articles 12 Footwear, headgear, etc. 13 Articles of stones 14 Precious stones, etc. 15 Base metals and prods 16 Machinery, etc. 17 Transport equipment 18 Precision instruments 19 Arms and ammunition 20 Miscellaneous manufs 21 Works of art, etc. Note: Source : Calculations exclude specific duties and include the ad valorem part for alternate and compound rates. Excluding in-quota rates. WTO Secretariat calculations, based on information provided by the Members.

13 13 Developing nations have participated in GATT/WTO Rounds for decades, but not on the basis of reciprocity. Until the 1970s, the GATT was a rich-man s club. Most poor nations were colonies or newly independent states, few of which wanted to join. When the merit of membership did become to be appreciated by some developing nations in the 1970s, the GATT adopted its Enabling Clause which should have been called the disabling clause since it disables most of the GATT s discipline for poor nations. The Enabling Clause meant that developing nations did not have to make tariff concessions in order to gain better market access. Or, more to the point, GATT Rounds did not turn developing country exporters from bystanders into free traders. Under these rules, the multilateral talks had no impact on the array of pro-trade and anti-trade political forces within developing nations. Little wonder, then, that they decided to stay with the initial level of their tariffs, which were, after all, politically optimal to begin with. In my view, the main role of the WTO is to help member governments fix the domestic politicaleconomy externality Olsen s asymmetry that prevents governments from doing what is good for their own nation, namely opening up to international trade. This fix was disabled for developing countries by the Enabling Clause. Not surprisingly, most developing nations trade policies were thus determined by what was politically optimal and this was what was economically optimal for politically well-organised special interest groups. In most cases, this included import-competing industries that had been created by infant-industry protection policies but had failed to grow up Political Economy Logic of Spreading Regionalism: the Domino theory Multilateral trade negotiations are not the only way tariffs have been lowered globally in the post war period. Indeed, at least among the rich nations of the world, regional liberalisation has probably been as important if not more important than multilateral tariff cutting, although the two are linked. Regional trade liberalisation does not occur randomly. The waves of regionalism that have occurred since the 1950s have followed a clear pattern a pattern that is called punctuated equilibrium in the world of evolutionary biology. Starting from a relatively stable situation in terms of trade barriers, some idiosyncratic shock to the system sparks a preferential liberalisation and this initial spark triggers a sequence of knock-on liberalisations. This is what I call the domino effect since it recalls how falling trade barriers in one set of countries tends to trigger falling barriers in other countries in a domino-like fashion. In the early 1990s, I developed an explanation for this pattern that I called the domino theory of regionalism. 5 This theory is simple. It starts with an idiosyncratic shock, such as deeper integration of an existing regional bloc or formation of a new FTA for political reasons. This triggers membership requests from countries that were previously happy to be non-members. Why would these governments change their minds? The stance of a country's government concerning membership in a regional bloc is the result of a political equilibrium that balances anti-membership and promembership forces. Among the pro-integration forces are firms that export to the regional bloc. Since closer integration within a bloc is detrimental to the profits of non-member firms (trade diversion and negative terms of trade effects), closer integration will stimulate the non-member exporters to engage in greater pro-integration political activity. If the government was previously close to indifferent (politically) to membership, the extra activity may tilt the balance and cause the country to join the bloc. If the bloc enlarges, the cost to the remaining non-members increases since they now face a cost disadvantage in an even greater number of markets. This second round effect will bring forth more pro-integration political activity in non-members and thus may lead to further enlargement of the bloc. The new political equilibrium will involve an enlarged regional trading bloc and the cycle repeats and regionalism may spread like wildfire (Baldwin 1997, p. 17). 5 See Baldwin (1993, 1997).

14 14 If the trade bloc is open to expansion, virtually all nations who depend heavily on the bloc will end up signing preferential trade agreements. If the bloc s natural enlargement 'burn-path' is barred, the new political economy flames may find vent in preferential arrangements among excluded nations. Notice that such regionalism could occur despite any progress being made in ongoing multilateral talks, unless these also promised to fully offset the discrimination. An old idea. The basic idea is not new. With hindsight, one can discern a version of it in Jacob Viner's account of how dozens of German principalities and city-states were cajoled and coerced into joining Prussia's Zollverein between 1819 and 1867 (Viner 1950 Chapter V.3). But the notion is probably much older. Bismarck himself probably understood the political-economy dynamics of trade diversion. Be that as it may, it is important to minutely examine the internal logic of even the oldest ideas and this requires maths. To my knowledge, Baldwin (1993), which deals with EU enlargement, was the first formal framework to explore the dynamic political-economy implications of trade diversion. Several subsequent studies - most notably Yi (1996), Ethier (1996) and Siedmann and Winter (1997) - have elaborated the basic framework. Sapir (2001) provides some empirical support for the domino theory. With this analytical framework for thinking about trade liberalisation in hand, we turn to the current situation in East Asia. 3. CURRENT STATE OF PLAY When it comes to East Asian regionalism, the state of play is easily summarised it is a mess. Dozens if not hundreds of trade deals are under discussion, under negotiation, or already signed. Even limiting the universe to the deals that have been signed or are near signing, it is clear that East Asian regionalism is marked what I dubbed the Noodle Bowl Syndrome. 6 As I wrote about Europe in Baldwin (1994) A web of trade deals can create a nightmarish tangle of administrative procedures that raise costs for enterprises and for governments. Most costly of all are those dealing with rules of origin." 7 Figure 8 makes the point graphically. The figure shows each FTA in the region that has been or is near to being signed. The ASEAN-Korea FTA (AKASEAN), ASEAN-China FTA (ACFTA) and ASEAN-Japan FTAs (AJFTA) are drawn as a separate deal for each bilateral relationship. Given the way ASEAN negotiated or is negotiating its FTAs with the three big economies in the region, the tariff charged on a particular product can be different for each of the 30 bilaterals. Likewise, all the bilateral trade links inside AFTA are listed as separate agreements given the disjointed nature of ASEAN. Due to ASEAN s rather unique method of preferential liberalisation, the degree of market access faced by an exporter of any particular product based in any particular ASEAN nation varies according to the ASEAN destination market concerned. 8 For example, we cannot view the Malaysia-Indonesia preferential tariff structure as identical, or even close to the Singapore-Philippines preferential tariff structure. Because of this, each bilateral relationship between the regional hubs (Japan, Korea and China) and the ASEAN spokes (Malaysia, Thailand, Indonesia, the Philippines, Singapore, Vietnam, Laos, Cambodia and Myanmar) is, de facto, a 6 I first used this term was in Baldwin (2004), but it should be thought of as the East Asian version of Bhagwati s famous spaghetti bowl problem. It is difficult to track down the exact spaghetti-bowl cite since even Bhagwati does not provide a references when he uses the term. The first reference I could find was a column by New York Times journalist Peter Passel, where he quotes Bhagwati as saying: "spaghetti bowl of tangled, inconsistent trade standards that just can't be good for efficiency." February 4, See the chapter entitled Hub and spoke bilateralism in Baldwin (1994); from 8 The basic point is that each ASEAN chooses its own preferential tariff rate on each good, including the ability to exclude or substantially delay liberalization of sensitive goods, but it is granted preferential treatment only on the goods it has granted such treatment.

15 15 unique pact (ASEAN has plans to substantially synchronise them sometime in the next decade). These agreements, and my assertions, are discussed in more detail below. Figure 8: The East Asian Noodle bowl syndrome. Noodle bowl syndrome HK China Korea Japan Myanmar Philippines Cambodia Thailand Indonesia Laos Vietnam Singapore Brunei Malaysia Note: The map shows : FTAs signed or under negotiation in January Source: Author s compilation The Noodle Bowl Simplified One branch of the Impressionist movement, Pointillism, consists of paintings made up of thousands of small coloured dots. Focusing on the dots provides the viewer with much information but little understanding. To get the big picture, one has to step back and ignore many of the details. A similar thing is true of East Asian regionalism. The situation in Figure 8 looks hopeless complex, but it becomes clearer when one ignores the FTAs that are inconsequential from the broader international trade perspective. For example: Singapore has signed a world-beating number of agreements, but these are irrelevant to broader questions of East Asian regionalism. Singapore is a small economy by world standards and two-thirds of its GDP is generated by its service sector, with trade-related services like shipping, insurance, finance and customs clearing services playing an important role. Moreover, the City-State long ago unilaterally lowered it tariffs to zero, so an FTA with Singapore either avoids substantial liberalisation or it is a unilateral concession by Singapore s counterparty. Consequently, the proliferation of Singapore s FTAs can be usefully omitted from Big-Picture thinking about East Asian regionalism and this despite that fact that Singapore accounts for about a tenth of intra-east Asian trade. Likewise, the trade deal between China and Hong Kong and China and Macao are no more relevant to the big picture than a free trade deal between Tokyo and Hokkaido would be. In particular, since Hong Kong maintains zero MFN tariffs and exports few goods to China that compete with exports from East Asian nations, its preferential deal with China is essentially irrelevant to East Asian regionalism. It is also useful to leave out the ASEANs free trade deal, AFTA, from the big picture for the simple reason that it not used. As I discuss and explain at length in Section 4.2, that vast majority of

16 16 intra-asean trade pays the MFN tariff i.e. the utilisation rate for AFTA is very low so we cannot honestly say that the 45 bilateral trade relationships among ASEANs are subject to free trade. Another important point is that most of the trade ties in Figure 8 cover trade volumes that are inconsequential. To make this point, Figure 9 shows all the East Asian FTAs that have been signed by January 2006 and cover at least 1% of intra-east Asian trade based on 2002 figures. (The thickness of the lines is proportional to the volume of trade.) All FTAs involving Singapore have been excluded for the reasons given above. The trade volume requirement immediately eliminates all of the intra-asean except the Thailand-Malaysia link and Singapore s link with the four biggest ASEAN traders, Malaysia, Thailand, Indonesia and the Philippines; the no-singapore condition eliminates these 4 bilaterals. When we have simplified the noodle bowl by only focusing on bilaterals that might be of consequence to third nations in the region, the picture is quite stark. We immediately see that real East Asian regionalism is by my definition marked by a hub-and-spoke arrangement with two hubs, Japan and China. It is important to note that many of the negative aspects of hub-and-spoke regionalism are mitigated by AFTA, but from a systemic point of view, the key trade arrangements in East Asia i.e. the trade deals that cause concern to third nations are dominated by China and Japan. This is what I called the East Asian bicycle pattern in Baldwin (2002). Figure 9: Noodle Bowl Simplified. Noodle bowl simplified, 2005: FTAs where bilateral trade is > 1% intra-east Asian trade* HK & Macao China Korea Japan Myanmar Cambodia Philippines Thailand Indonesia Laos Vietnam Singapore Brunei Malaysia East Asian FTAs signed by 1 January 2006 that cover bilateral trade that exceeds 1% of total East Asian trade. * Singapore FTAs excluded. By focusing on the trade ties that really matter from a political economy perspective, we can more easily organise our thinking about how East Asian regionalism is likely to evolve in the medium run and how it should best be managed.

17 17 I turn now to detailing the main elements China-ASEAN China s unexpected offer in 2000 to sign an FTA with the ASEANs triggered a domino effect in the region, as I explain at length in Section 4. The effect, however, occurred in slow motion. It was only after years of difficult discussions that China and the ASEANs signed an FTA in December This agreement is complex and affects East Asian regionalism in complex ways. Although it is a single agreement in principle, it functions as a set of bilaterals due its novel form of conditional reciprocity. The FTA allows each of the 11 parties (10 ASEANs and China) to unilaterally choose the goods it wants to put into the normal liberalisation track ( Normal Track ) and those it wants to liberalise more slowly or not at all (by placing the good in the Sensitive Track ). Treatment of the Normal Track goods is in line with standard multi-party FTAs practices. That is, preferential tariff cuts made to any party are automatically extended to all parties (Annex 1, paragraph 2). Preferential tariff cuts on Sensitive-Track goods, however, operate in a more complex manner. The basic effect is that a nation s exports of its Sensitive-Track goods face tariffs that are at least as high as its own tariffs on the product concern. 9 For example, Malaysia s high tariff on cars means that its car makers will face high tariffs when exporting to the 10 other partners, with the actual rate being the higher of two rates Malaysia s own preferential tariff on cars and the preferential tariff that each of the other 10 partners charges on cars. Although it was signed in December 2004, has not been notified to the WTO as of November 2005 (most, but not all of the 11 participating nations are WTO members). Thus although I cannot be sure, it appears that China-ASEAN will be notified to the WTO under the Enabling Clause, so it will be subject to almost no WTO discipline Japan-ASEAN bilaterals Japan proposed an FTA with the ASEANs in January 2002 and the parties signed a Joint Declaration in November 2002 to establish closer economic relationships that could include FTA. In October 2003 they adopted a Framework Agreement laying out further steps, but little progress has been made as of early Japan s bilateralism strategy Japan has instead embraced bilateralism negotiating bilateral deals with those ASEANs where the mercantile interest is greatest on both sides, namely Malaysia, Thailand, Indonesia and the Philippines. The negotiating Framework Agreements were signed in 2003 and talks started in early Only one of the four agreements was complete by January 2006, but all four are likely to be finalised in 2006 if nothing else, to prevent Malaysian firms from gaining too great of an edge over their AFTA partners in the large Japanese market (domino effect). The Japan-Malaysia FTA (JMFTA), signed in December 2005, sets an extremely important precedence for East Asian regionalism. Indeed, the deal reveals what will surely become the basic pattern for FTAs between the industrialised East Asian nations Japan and Korea and the rapidly 9 Paragraph 6 of Annex 2: The reciprocal tariff rate treatment of tariff lines placed by a Party in the Sensitive Track shall be governed by the following conditions: (i) the tariff rate for a tariff line placed by a Party in the Sensitive Track must be at 10% or below in order for that Party to enjoy reciprocity; (ii) the reciprocal tariff rate to be applied to a tariff line placed by a Party in the Sensitive Track shall be either the tariff rate of that Party s tariff line, or the Normal Track tariff rate of the same tariff line of the other Party or Parties from whom reciprocity is sought, whichever is higher; and (iii) the reciprocal tariff rate to be applied to a tariff line placed by a Party in the Sensitive Track shall in no case exceed the applied MFN rate of the same tariff line of the Party or Parties from whom reciprocity is sought.

18 18 industrialising East Asian nations such as Indonesia, Thailand, Malaysia, the Philippines and Vietnam. Due to extremely well-entrenched special interest groups, Japan s government finds it impossible to liberalise most agricultural goods, but especially rice and beef. Japan s FTA partners use this impossibility as a lever to win concessions from Japan on their favourite importsubstitution industrial sector. In the case of Malaysia, the favoured sector is automobiles. However, in the JMFTA the Malaysian government has been very farsighted by agreeing to phase out tariffs but insisting that Japan provide technical assistance for its car industry. Presumably the idea is that auto tariffs will in any case have to be phased out eventually and this JMFTA provides an opportunity for boosting Malaysian auto industry competitiveness in advance. It is also possible that Malaysia felt compelled to make concessions on autos since it knew Japan was also talking about cars with its rivals in Indonesia and Thailand and ASEAN did not coordinate the four sets bilateral talks with Japan. The Japan-Thailand FTA was agreed in principle by the two heads of government in September 2005, but a finalised document has yet to emerge as of early January The agreement in principle, however, shows the same pattern as the JMFTA. Japan gets to exclude rice and other sensitive agricultural goods and in exchange, Thailand gets to keep high tariffs on autos Korea-ASEAN Korea whose market is far less important to ASEAN exporters than China s or Japan s is playing catch-up in the FTA department. The FTA processes between the ASEANs on one hand and China and Japan on the other formally started with Framework Agreements in 2002 and 2003 respectively. Korea s first step, a Joint Declaration, came a year later in October 2003 and its Framework Agreement with ASEAN was signed in 2005 and even then difficulties between Thailand and Korea prevented publication of the AKFTA framework. The differences between Korea s and Japan s experience with the ASEANs is extremely revealing and helps illustrate the political economy forces that tend to create hub and spoke trade arrangements. For both Korea and Japan, rice is an extremely sensitive issue in domestic politics, while rice is an important export for several ASEANs but especially for Thailand. When the rice issue came up in the Japan-Thai FTA talks, Japan s large market high level of imports from Thailand gave it the bargaining leverage necessary to exclude rice from Japan-Thailand liberalisation. Something like 15% of all Thai exports go to Japan, so the market-access gains to Thai manufactures exporters were a prize that compensated the failure to gain better access for Thai rice. Simultaneously, the large size of the Japanese auto industry made Thai car producer believe that it was essential to maintain their protection against Japan. Thus the fact that Japan was so large compared to Thailand was critical to establishing the economic basis of the political deal, rice for autos. Moreover, Japanese FDI has been critical in Thai industrialisation over the past two decades. Korea, by contrast, was not able to get a deal with ASEAN as a whole in December 2005 since Thailand insisted that Korea could not exclude rice from its preferential market opening. Why would Thailand let Japan do something that it would not allow Korea to do? Korea absorbs only 2% of Thai exports and Korean FDI has not played as significant a role in Thailand s development. Moreover, while the Japan-Thai trade relationship is lopsided (Thailand cares far more about the Japanese market than Japan does about the Thai market), the Korea-Thai relationship is more balance, and balanced at a very low level neither nation depends upon the others market for more than a percentage point or two of its exports. Because neither Korean nor Thai exporters care a great deal about each others markets, the protectionist tend to win when it comes to the Korea-Thailand link (low exclusion index), but tend to lose when it comes to the Japan-Thailand link (high exclusion index, i.e. hub-spoke).