MERCOSUR'S OPEN REGIONALISM AND REGULATION: THE ROLE OF BUSINESS ACTORS. FOCUSING ON THE PETROCHEMICALS AND STEEL SECTORS'

Regionalism, in the 1990s, appears worldwide to be an outward oriented phenomenon of political economy (Mittelman, 1996:191, 205), and a permanent feature affecting global trade (Hugues Hallet et aI., 1994; De Me10 et aI., 1993). Nonetheless, when it comes to the case of Mercosur l formed by the states of Argentina, Brazil, Paraguay and Uruguay. The effects of such policy-driven efforts and free trade stand have been seriously questioned (Yeats, 1997)2 on the basis of weak business-state coalitions and protectionist business strategies, remnants from the lSI era (Pastor and Wise, 1995). For some others like Payne and Bartell (Payne et aI., 1995:257) or Bouzas (1997:14-15), Latin American business leaders behave differently from the past, as reflected for example through the indicator of intra-regional Mercosur investment flows. These contradictory assessments of Mercosur as well as of business actors' behavior regarding regionalism beg for some clarification. To what extent did business people, the other key actors besides government officials, pursue the path of regionalism in Mercosur? To what extent and in which manner were they associated with the regulatory process involved in regionalism? If they were, what reasons led business actors to engage in the regional policy making in the first place? The argument of the paper is as follows: In its very conceptual tenets, the 1990s Mercosur's "open regionalism" has sought not only the support of the business actors, but also their active involvement. This involvement has indeed taken place, giving rise to a new regional policy dynamics. Yet, business involvement has also been a response to their autonomous perceptions of, and interests within, regionalism.


II. THEORETICAL AND METHODOLOGICAL CONSIDERATIONS
Argentina and Brazil, which initiated a bilateral attempt at regionalism in 1986 (the Argentine-Brazilian Integration Program -ABEIP), form the axis of Mercosur in economic 3 as well as in political terms. Therefore, this paper concentrates on this axis, while it focuses on the regional strategies of business actors belonging to two industrial sectors, namely petrochemical and steel. Since the volume and diversity of actors logically increase in any process of regionalism, desegregating into sectors can improve the analysis of the 'meso level' -of the intermediation of the actors' interests at the transnational level (McAleavey, 1994). Studying actors' political strategies in regionalism through sectors shifts research from macro diversity at the national level to responsible politics where actors become more clearly accountable, and takes into account specific interests that are common to certain groups and sectors (Booth, 1993: 49-53). Already in the 1980s, Strange and Toozen (1981: 17) had called for analytical desegregation that could lead to a reconstruction of theoretical statements based on "analysis by economic sector rather than on a macro level". In the twentieth century, "the modem industrial enterprise ... dynamics produced its three most significant historical attributes. First, such enterprises clustered from the start in industries having similar characteristics. This is a main reason why sectors' studies are particularly interesting. Second, they appeared quite suddenly in the last quarter of the nineteenth century. Finally, all were born and then continued to grow in much the same manner" (Chandler, 1990:18). Thus, many economic indicators concerning business strategies 4 are sector-bound (Chudnovsky, 1993:4), and a sectoral analysis is essential to fully comprehend business actors' behavior.
According to Milner, industrial sectors with increasing returns to scale of production "should be the ones who pursue a regional strategy" (Milner, 1996:80), and thus, would be more likely involved in regionalism. Industries that are not under perfect competitive conditions would prefer to export than to invest abroad. A customs union would correspond to the firms' interests of maximizing profits, through trade suppression and cost reduction combined with product differentiation, particularly when the home market is small compared to the most efficient scale of production. It would also provide them stable market access. These considerations apply to both petrochemicals and steel, where most products are commodities, i.e. standardized products. thus characterized by increasing economies of scale. Their plants are technically indivisible and, given their increasing dimensions pushed by large economies of scales, they require enormous capital investment. The fact of being commodities means that the bulk of competition lies on pricing. Vulnerable to changes in global and national economic conditions, the high costs of capital and of halting the plants' operation make any altemative preferable to temporarily halting production or closing down. Both the cost of capital and the certainty over the 'mles of the game' or policy regime are important for their long-term strategies. As a result, chemicals and petrochemicals (together with oil) tend to form oligopolies and have been the first most internationalized industrial sectors 5 . In steel, global firms have also slowly appeared in the 1980s 6 , and in the 1990s, steel firms that remain dependent on their home market and without alliances abroad are perceived as non-competitive (Hudson, 1996:160). Accordingly, most steel and petrochemical firms from the Mercosur region are oligopolies on their way to intemationalization (see Peres Nunez, 1993), and belong to the main national economic groups or holdings7.
These two sectors have a strong impact on their national economies, since they are important in terms of capital investment as well as in terms of their active role within the domestic industrial networks. As producers of intermediate goods, they are central to the economy due first to their forward and backward linkages, and second to the fact that they shape national prices and competitiveness in terms of average costs for the rest of the economy. As a result of their special production cycles and their capacity to influence the rate of inflation, governments may target them for special treatment. All these characteristics make them key business actors, likely to internationalize and to be heard by governments. However, this is certainly necessary, but not sufficient, to explain their involvement in Mercosur regionalism as a regulatory and policy-making process. Within Mercosur, economists have studied how these sectors have reacted to their changing national economic contexts 8 . Simultaneously, other authors have studied various actors' reactions to regionalism, including business in general (Hirst, 1991, Hirst et aI., 1993Alvarez, 1995). Nonetheless, the sectoral dimension of business actors' political and social role in policy making have yet to receive the attention it deserves (Schvarzer, 1993:380, 386). This paper is meant to be a modest contribution in this regard. The discussion has been grounded on 74 personal interviews conducted in 1995 in Argentina and Brazil, and later analyzed within Atlasti qualitative software, covering as much as possible the 'power map' of the main business actors from both petrochemicals and steel involved in Mercosur policy-making. Business actors include owners and managers of enterprises and holdings, as well as representatives of trade and/or industrial chambers, be they national, bilateral or Latin American.

Mercosur's Regulatory Process
Mercosur's objectives and status have been set out in its founding treaty, the Asuncion Treaty of 1991, as well as in the Protocol of Ouro Preto of 1994. It intends to create a common market with a common policy towards third parties, embodied in its common external tariff, and a free area for the movement of goods and services, capital and labor. It also aims at macroeconomic coordination as well as at harmonizing sectoral policies. As regards to its status, Mercosur has become both a customs union and a legal person of international law as from January 1995. By the latter feature, Mercosur officials are able to negotiate as a unit, obtaining international recognition and a greater bargaining power in international negotiations. The customs union, though, is being gradually implemented, and is expected to cover all sectors ofthe economy only in 2006 9 . Mercosur appears largely as a consequence of a decade long unilateralism lO • As seen earlier, Argentina started unilateral liberalization in 1987 and has been by far the most unmitigated case, while Brazil has followed a similar path since 1988, albeit more slowly. Thus, Mercosur was established amidst an ongoing and general reduction of trade barriers. In addition, these countries have a multilateral trade insertion in the world economy, their total exports roughly divided in similar percentages among North America, South America and the European Union 1 I. Their diversified articulation explains Mercosur's interests in a solid multilateral system under the surveillance of the WTO. Indeed, a strong' WTO may enable the defense of their interests without being excessively dependent on just one country or region.
In the context of liberalization in the two countries, rapid technological change and world restructuring in industrial sectors l2 , the first priority in the region was to attract capital flows and to provide a stable framework for capital integration. The two sectors examined here, above all petrochemicals, were particularly concerned with both the issue of capital fOlmation and the question of scales 13. Indeed, regionalism for those sectors was connected to the goals of increasing scales, enabling their holdings to expand and to connect their firms into regional chains of production and distribution. In such a way, the probability of trade conflict could diminish while that of "capital formation,,14 could increase. Moreover, there was a governmental recognition that, if this failed, regionalism would fail as we1l 15 . Correspondingly, a primary goal of both national governments has been that of macroeconomic stability. For the governments, "the question is to create the conditions for attraction of capital" (Campbell, 1994).
In this context, the ministries of economy and finance acquired most of the power in decision making in both countries, while foreign ministries were subordinated to the economic agendas. The key actors have been government officials and business representatives (owners and managers of enterprises and holdings, as well as representatives of trade and/or industrial chambers, be they national, bilateral or Latin American ones). Mercosur has so far remained strictly intergovernmental and based on the Argentina-Brazil bilateral axis. Its intergovernmental working bodies have neither the autonomy to set their agenda, nor financial resources of their own. Moreover, their decisions can only be reached by consensus and cannot be implemented straightforward. They first need to be 'internalized' through state approval and endorsement. The only supranational regulation until 1995 was the Annex I of the Asuncion treaty, with the automatic implementation of decreasing tariff rates.
Mercosur's structure is minimal, with almost no bureaucracy of its own, with the exception of the Administrative Secretariat in Montevideo 16. Mercosur officials are also national civil servants, while the presidency of Mercosur rotates every six months among the presidents of the member countries. A limited number of decision-makers with rather stable positions have facilitated consensual decision-making, which in tum has enhanced personal channels 17. The definite Mercosur structure was established by the Ouro Preto Protocol, which came into force on November 1, 1995. The Protocol added the Trade Commission and the Economic and Social Consultative Forum. The former has a supervisory role and handles the complaints on trade matters, while the Joint Parliamentary Commission and the Economic and Social Forum have remained so far consultative. The current structure comprises six main bodies. In decision-making, these are the Common Market Council, the Common Market Group and the Trade Commission. The Joint Parliamentary Commission is the body of parliamentary representation. The two other bodies are the Economic and Social Forum and the Administrative Secretariat.
Ten working groups (WG), established in 1992, reviewed economic issues under the Common Market Group authority. The WGs had an important role in the so-called transition period (between 1991 and 1995) in which they were considered as ad hoc groups. Since 1995, the WGs continued to cover various fields. The task of these working groups has been to harmonize the national norms with the ideal of setting a "leveled playing field,,18. Mercosur's goal of harmonizing sectoral policies includes the issue of industrial policy instruments (e.g. concerning trade, tax and credit instruments). However, the issue of industrial policy in itself has neither been mentioned nor its possible instruments delimited, with the exception of the reference to sectoral accords before 1995, that will be explained later (Markwald and Machado, 1999:74;Leipziger et al., 1997:14). The restructuring has been left to the firms themselves.
In spite of a clear commitment to general liberalization, the sectorbased option was kept open within Mercosur through the so-called "Sectoral Accords" (SA). The aim was to restructure a sector across different states by pooling them together in the regional market. SA were meant to involve 'sensitive' 19 and capital-intensive sectors into regionalism. Due to the important structural asymmetries in the industrial sectors between the two countries 20 , the idea was to prevent a crisis that could be serious enough to stall the process. SA depended on voluntary action, and need not include all Mercosur countries. They were to be negotiated by the sector actors themselves, but their ratification was dependent on each state's approval and on the consensus among the countries involved. At the same time, they were supposed to increase the intra-sectorbased trade, counteracting the inter-sector~based emphasis of the general approach (Guadagni, 1993). The only SA ever approved by Mercosur, and included in the ALADI system, was the one of the steel sector. The reasons for this outcome will be argued in the next section on business involvement in regionalism.

Mercosur's Trade Effects
Regarding the trade effects in Mercosur, there has been a debate over the static ones. There are those who criticize them (Yeats, 1997), those who see them as a natural process resulting from shortened geographical distances (Machado and Markwald, 1995), and those who are enthusiastic about them (Irela, 1997, Leipziger et al., 1997. Mercosur trade growth has indeed been significant. Intra-regional exports increased by almost five times between 1990 and 1997, and grew from less than 10 percent of total Mercosur's exports in 1990 to almost 25 percent in 1997. 21 It is the issue of dynamic trade effects, however, that has been at the forefront of both state and business actors' consideration. According to Leipziger, Fritchtak, Kharas and Normand, the arguments for dynamic trade effects in Mercosur are three. These would stem from: 1) he capacity of larger markets to attract FDI; 2) from the exploitation of economies of scales larger than the present national ones; and 3) from the idea that the "domestic corporations first need to be toughened through exposure to limited regional competition" (Leipziger et al., 1997:6). As regards FDI flows, "between. 1994 and 1997 FDI growth in the Mercosur region exceeded 52 percent annually. Argentina and Brazil were the chosen destinations for more than 95 percent of all FDI entering the bloc. From 1990 to 1994, Argentina clearly exceeded Brazil in FDI competition, which underscores the importance of economic stability as a factor in attracting this type of capital. In 1995, one year after implementing the Real Plan in Brazil, the distribution of investments flowing into the two principle subregional members balanced out" (Mercosur Report No.4,. The growth of intra-industry trade is one of the three best indirect measures of dynamic effects from trade liberalization 22 (the other two measures being the reduction of price-cost margins in the tradable sectors and the positive growth of total factor productivityi 3 . Such measures point to the degree of rationalization and specialization taking place within the region, as well as of enlargement of scale economies. Machado and Markwald (1997) affirm that more than 60 percent of intra-Mercosur trade in manufactured products are exchanged through intra-industry trade.
Dynamic effects are translated into reality as Mercosur ensures effective market access to the firms of the member countries. Of the total universe of goods, around 85 percent have been liberalized from January 1995. Mercosur has an average CET of around 12 percene 4 , which is rather low compared to the region's previous levels of protection 25 • A handful of sectors are still protected (e.g. automobiles and sugar) and a few others will harmonize tariffs by raising Argentine ones (telecommunications, computers and capital goods). In the case of petrochemicals, most products enjoy a zero tariff, while iron and steel industry products have been included in a third type of mechanism regulating regional tariffs, called 'Regime of Adjustment'. This mechanism is meant to phase out the remaining barriers to trade, made of decreasing tariffs rates covering particular items until 1999. In the case of steel, the tariff range goes from 10 to 15 percent. Thus, non-tariff barriers have remained significant for many steel products (cf. Table 3, in Machado et aI., 1997: I 86).
The minimalist treatment applied to industrial policy at the regional level has allowed for national policy instruments such as anti-dumping and safeguards, mainly used by the smaller market of Argentina, principally in the case of steel. Later the safeguards for steel have been internalized into the Regime of Adjustment. In 1995, in petrochemicals, there was "one exception, included in the Regime of Adjustment. Polyethylene of high density enters with quota and then with tariff' (FWI, 53:19)*. In 1999, the situation remains the same. "Very few petrochemical products come under * FWI stands for 'Fieldwork Interview'. Interviews were concentrated on states and films' actors during the period under study, since these were the main actors in regionalism. In this paper, quotations from fieldwork interviews appear with two numbers separated by a semicolon. The first number indicates the interview document and the second one shows the number of the quotation within the interview itself.
Mercosur's adjustment regime, and practically all of them are traded with zero tariff' (Hasenclever and Lopez, 1998).
There are other barriers that remain due to the states' lesser capabilities to deal with them. One area concerns the uncertain and complex implementation of regional regulation and control of norms of origin. This affects more the petrochemical sector than the steel one, due to threats of triangulation through the smaller countries and the broader range of products involved. Another problematic area, beyond the sectors, is the difficult relationship between the goal of macroeconomic and exchange rate stability, with the increasing deficits in the balance of trade and the volatility in capital flows. After each main financial crisis (Mexico, Asian, etc.), the CET and interests rates are increased. Additionally, there is a tradition of using trade policy to solve trade deficits. In 1995 and 1996, Brazil exempted several products from the CET (agreed upon just a couple of months before) for the reason of maintaining price stability after the Real Plan. The first exemptions included petrochemicals and meant a reduction in the external national tariff to just 2 percent. Following exceptions included among others steel and fertilizers. As a tacit rule, the priority of stability means that, at the Mercosur level, uncertainty and a low degree of state compliance is part of regionalism. "In fact, the remaining obstacles for the full consolidation of Mercosur as a customs union appear to be less associated to the trade regime related to third countries, than to the instability in the levels of protection and other obstacles -of regulatory nature, bureaucratic and infra-structural -to the intra-regional trade flows" (Machado et aI., 1997:183).
These obstacles have raised many trade complaints from the private sector, which are handled first at the national level or by the Trade Commission. Legal disputes in Mercosur are managed first through political governmental negotiations, followed by ad-hoc arbitrage commissions 26 . Brazil in particular has been adamant against any supranational tribunal. The situation, which does not create legal jurisprudence, is open to repeated costly disputes over the same issues with the risk of differing interpretations and legal sentences. An emerging consensus is favoring the establishment of a private system of arbitrage 27 •

Mercosur's Open Regionalism
Mercosur's vision and set of values have been rationalized through the concept of "open regionalism". Essentially related to economic and political liberalism 28 , they include the search for stability, consensus, and a state with a minimal regulatory role. Evidence stems from documents, public presentations and personal interviews. In an Argentine document of the ministry of foreign relations, the rationale of Mercosur was defined as the following: "[Mercosur is] ... open regionalism, in the framework of market oriented national policies and adhesion to multilateralism ... In case of success, Mercosur constitutes a tool intended to improve business profitability and the expected returns from investment projects" (Campbell, 1994). Formally, the concept has been legitimized through Mercosur's relation with the EU, through their 1995 Framework for Cooperation that seeks to create a transatlantic free trade area. In that document, the governments affirmed "the importance of an open regionalism". While the concept may diminish any concerns about a competition between the USA and the EU for 'spheres of influence', the Mercosur governments wanted to signal their choice for a non-discriminatory bloc against outsiders. Mercosur has largely been in consonance with this stated vision of 'open' regionalism. The Asuncion Treaty had from the start a clause for enlargement. In addition, Mercosur has been placed within the ALADI system, which means that it is open to all Latin American countries. In December 1995, the Agenda 2000 reinforced enlargement as one of the main priorities for Mercosur, together with the goal of a common trade policy. Two countries have become associates: Chile and Bolivia. Canada and Central America have signed agreements with Mercosur. Negotiations to free trade are under way with the Andean Group as well as with the European Union. Finally, the Mercosur countries are participating within the Western Hemisphere Free Trade negotiations.
In public presentations given at the first meeting of the European Commission and Mercosur Joint Cooperation Committee (Brussels, June 11-12, 1996), the regional value consensus underlying Mercosur's regionalism was laid open. In that meeting, business actors declared that "the Mercosur countries have shared three general features: a representative system of government, neo-liberal economic policies, and the diffusion of military threats. These features were related to the priorities given first and foremost to stability and the attraction of capital investment, and they were the result of a power consensus among the elites,,29. Such a consensus focused on "four issues: macro policies of adjustment in terms of stabilization, micro economic policies in terms of deregulation, trade policies in terms of liberalization, and the private sector as an engine of capital accumu1ation,,30. In other words, affirmed Brazilian Celso Lafer at the Brussels Seminar, "the key issue for Mercosur is competitive insertion into the world economy".
The UN Economic Commission for Latin America and the Caribbean (ECLAC) has provided the rationalization of the concept for the whole of Latin America. In its publication "Open Regionalism in Latin America and the Caribbean" (ECLAC, 1994), "open regionalism" is used to "refer to a new process that results from reconciling the two phenomena ... the interdependence that stems from special, preferential agreements, and that which basically arises from the market signals that are produced by trade liberalization in general" (ibid.: 12). "Open regionalism" has been conceptualized as a strategy to ensure the insertion of LDCs into a world conceived as multipolar, while preventing their turning into a closed trading bloc. To comply with the concept, several conditions are to be met (see Fuentes, 1994), the main ones being trade liberalization, harmonization of standards, and flexibility in the institutional regional regimes.
Business actors were invited to participate on four grounds. The first ground was "to avoid the dangers of premature institutionalisation", that "owing to the absence of a single authority with a monopoly on power" at the "multilateral level" (ECLAC, 1994:97), could spur intense conflicts of interests. Second, regionalism had to reflect the regional trends towards democratization and representative government. Thus, and third, by giving voice to new actors, regional institutions could respond better to the aspirations of various social groups. Fourth, the de facto integration phenomenon -of firms and investment flows -had to be included. The latter "largely conditions the evolution of policy-driven integration, especially when it occurs in a context of globalization, liberalization and deregulation; it also requires more active participation by non-governmental actors, such as business, labor and cultural groups, in the design and implementation of integration policies (ECLAC, 1994:97-98).
The late director Faynzylber clarified ECLAC's thought in an interview (Fajnzylber, 1994). Fajnzylber maintained that ECLAC sees an "open participatory system" as intrinsic to its proposal, and expected that on a corporate basis the "principal actors that take parLin the production process" (Ibid:207). "Open regionalism", in its very tenets, has sought the active political involvement of business actors in the state-originated strategy of regionalism, and not only their support, as will be seen in the next section. As a result, there have been four main policy implications stemming from "open regionalism". First, trade is expected to increase interdependence and to be conducive to regionalism through "spillover" effects. Second, there would be a neutral process in policy selection, the best of which would stand out by itself. Third, there would be no industrial policy or regulation as such. Finally, business actors should actively participate in the decision-making, design and implementation of policy.

IV. BUSINESS ACTORS' RESPONSE TO MER CO SUR REGIONALISM
The previous section analyzed Mercosur's "open regionalism". This one reviews how business actors have responded to it. Have they been associated with the regulatory process involved? Given the region's structural heterogeneity (Becker et aI., 1992:140), the analysis of business actors' behavior is complex. In general, this issue has been debated in terms of 'weak versus strong' actors or coalitions (see Payne et. aI., 1995). However, the main issue at stake here is whether business' involvement has continued with their traditional fragmented and particularist behavior, or whether they have been able to accept general rules of the game and a general articulation of the political arena and the policy-making process (Viguera, 1995).
Regionalism implies a policy making process that covers a regional space with the purpose of creating sustainable regional patterned relations (sociopolitical as well as economic). It is, therefore, a politico-economic process entailing power relations, which should not be prejudged as 'policy-led' by the state. This is where a meso-level research could be promising. Studies on regionalism have shown that it entails at least a minimal process of institutionalization, which takes place as emerging patterns of relations can be coordinated in a regular fashion, embodying particular values and normative rules that become approved, regularized and possibly formalized into new structures. As regionalism becomes institutionalized, perceptions tend to shift and to consider it as a long-term policy in the countries concerned.
In thel980s, the debt and the state's fiscal crisis affected the two sectors, which had evolved during the ISI* period. Thus, they experienced national economic crises, uncertainty and instability. In the 1990s, they underwent structural national reforms. First in Argentina (1991Argentina ( -1995, and later in Brazil (1994Brazil ( -1998, the two sectors were privatized, their tariffs reduced, and their regulation dismantled or diminished 3 !. Although it is difficult to strictly separate business' actors strategies linked to internal reform from those due to regionalism, this section will emphasize the latter. Attentive to Mercosur's very beginning, business representatives had been preparing the ground for regionalism. Parallel to their recognition that regionalism was from the start a strategy led by the states, already in the mid-1980s (when the Argentina-Brazil Integration Program started) some of the sectoral chambers from both countries had begun to meet. They discussed the regional situation and possible strategies, perceiving a new era marked by globalization and internationalization 32 . Already at the start of Mercosur, the Bilateral Argentine-Brazilian Trade Chamber was playing a significant role 33 • The two sectors were experiencing considerable changes in behavior 34 • In petrochemicals, "transformations in organization, quality, * Import Substitution Industrialization processes, etc., not seem attributable to Mercosur's per se, but to the trade liberalization and deregulation carried out since the start of this decade. However, the establishment of Mercosur did generate new business strategies, including the installation of commercial offices in the partner country, the formulation of complementarity schemes between the affiliates of transnational companies, direct investments in the neighboring country, partnerships between Argentina and Brazilian firms, etc." (Hasenclever and Lopez, 1998). Similarly, the portrait of the steel sector at the Word Bank was that "the big [Mercosur] firms have entered into agreements of co-operation, while the [Brazilian] medium scale ones have tried to enter into the Argentinean market through the control of distribution and/or associating with the few remaining independent laminators. The latter have been the most marginalized from negotiations" (interview with Dr. Meo, Washington, April 1997).

The New Competition
The privatization process, one of the main components of structural refom1, has had two special impacts as far as business actors' involvement in regionalism is concemed. First, it has allowed regional business alliances to take place, increasing the possibility of building new business strategies and rationale to match the new environment; at the same time, it has allowed a restructuring of the sectoral chambers in the region, providing the latter with new roles and capacities. Second, it has provided business actors with the means to design the Mercosur the sectoral accords (SA) with the goal of encompassing the whole process. In similar sequence, examples of network building are considered first, followed by a discussion on the SA.
Along the privatization of the sectors, some important long-term regional alliances took place. Those engaged were considered strategically offensive-oriented holdings and 'first-movers' (FWI 19:4). In their negotiations, sectoral chambers had an instrumental role to play, offering a restructured neutral place to meet and coordinate. The Latin American chambers of ILAF A (Latin American Steel Institute) and APLA (Latin American Petrochemical Association) in particular took up the latter role 35 . In November 1992, the Argentine holding Techint purchased Somisa, that country's main state steel plant, together with Usiminas, (Brazil), CVRD (Brazil), Acindar (Argentina) and Aceros del Pacifico (Chile). Usiminas bought 5 percent of Somisa, integrating Siderar's board. This was Usiminas' first investment abroad, which afterwards multiplied in an expanding network on its own. CVRD also participated with another 5 percent, locking in the 10ng-tem1 provision of raw materials in special conditions and prices (CVRD. privatized in 1997, is the regional supplier of iron ore and a world price setter). Acindar soon left the venture. In July 1993, Somisa was transformed into Siderar, specializing in the production on flat products while discontinuing the non-flat ones, focusing on improving marketing and servicing the internal market. In 1998, "Siderar is part of Siderurgia Amazonia Ltd, a consortium that owns 70 percent of formerly state-run Venezuelan steelmaker Siderurgica del Orinoco" (Reuters, February 4, 1999). The latter is now partly own by Techint. Meanwhile Usiminas is enlarged in 1999, merging with Cosipa of Sao Paulo but maintaining its name. Siderar, Usiminas and Cosipa all produce flat steel products destined in a large part to the automotive industry.
The rationale of the first alliance was not to control Somisa. This strategy has been considered as "soft regional integration" (interview with a consultant from Booz, Allen and Hamilton) for the purposes of exchange of information and other types of cooperation, including technology, training and experts' exchanges, while preventing potential actions of dumping (F-W I 22:5). Techint and Usiminas interviewees clarified: it was "to bring up new opportunities". As an example, they would supply at least three of the main automobile companies operating within Mercosur (Usiminas manager, FWI 70:2). In this case, privatization opportunities are used to build a vertical alliance from raw materials to automobiles, also providing services and credit to the clients. The rationale was the creation of a strategic regional network. On that ground, further alliances took place between Techint and Brazilian holdings (FWI 49:8, 4:6, 22:7, and 23:27, also Argentine Embassy in Brazil, 1995), showing the vast array of moves in which regional holdings may be involved (cross-shareholding, joint ventures to establish new firms, commercialization/distribution agreements, and regional supply to Mercosur infrastructure projects, among others).
The common content to all has been the accomplishment of longterm cooperative networks, a goal rationalized under the concept of 'business scales', as discussed in the last section.
The petrochemical sector has followed a similar pattern. In the petrochemical sector, Argentinean Perez Companc chose to strengthen itself through the firm PASA, already present in Argentina. With it, Perez Companc came to control the refinery of San Lorenzo (where PASA had 15 percent and Perez Companc 33 percent of the shares, thus obtaining the majority), and acquired voice and voting powers in Unistar (a joint venture with Monsanto), opening the possibility to position itself in the sub-sector of fertilizers. In the 1990s, Perez Companc had engaged in high value added production for automotive use and fertilizers for agriculture. In this case, the holding, whose core interests lay in oil, purchased and transformed one firm, and "redefined it as its petrochemical arm" (Argentine petrochemical consultant, FWI 36: 11). This was not only a case of vertical integration from oil into petrochemicals. The "arm" would be the coordinating center to expand including the building of a network in the latter.
Perez Companc and Odebrecht have had two important regional joint ventures 3G • Moreover, they bid together for the Bahia Blanca's petrochemical complex, the major petrochemical privatization in Argentina. In this last privatization case, two were the major competitors: Perez Companc in alliance with Odebrecht (through its subsidiary Copesul), and YPF in alliance with Dow Chemical. The latter won, triggering a 'counteroffensive' from Brazil (interview with Odebrecht manager).
By 1997, there were two major Argentinean oil holdings, both vertically integrated into the petrochemical sector and both highly professionally managed and internationalized, YPF and Perez Companc. In 1997, YPF, Dow and Perez Companc owned 75 percent of the sector in Argentina and all the internationally competitive firms (Lopez, 1997: 355). In 1999, though, as the Argentine State withdraws further from YPF, selling its last shares, the situation may change again. Spanish Repsol, already participating in YPF, may take over its control, likely preserving the holding from any financial crisis in the region.
In Brazil, Petrobras, in a redefined institutional situation since 1997, is re-engaging in the petrochemical sector 37 , to create a renewed core in the industry. Odebrecht is a major conglomerate producing resins, polyolefinic specialities and PVC among other products. Entering the petrochemical sector in 1979 through the Companhia Petroquimica Camacari, Odebrecht became a transnational in the 1980s 38 , and participated heavily in the 1990s privatization process in Brazil. Now it is the most important holding in petrochemicals, after obtaining participation in all petrochemical complexes in Brazil, and is the Brazilian leader in polyethylene and polypropylene. Odebrecht installed a representative sister-company in Argentina to evaluate the market and future company acquisitions and/or joint ventures. The objective was to create a business base for Mercosur, including Chile. As seen above, it expected to enter into Argentina's Bahia Blanca complex. Since YPF and Dow Chemical won the Bahia Blanca's petrochemical privatization, Odebrecht pursued its own expansion and a broader alliance with the oil sector (Jornal do Brazil, November 3, 1998).
The establishment of joint ventures provides a regional supply base for sectors such as the regional automotive industry and the new Mercosur infrastructure, enabling the formerly national, now regional, holdings to engage in the new type of competition. Moreover, it is clear that the concentration of each industry and the building of sectoral networks do not stop at the Mercosur level. They tend to enlarge beyond, as the strategy gradually extends into Chile, Venezuela, Peru, Ecuador and Mexico in Latin America, and, beyond, is introduced into China, Southern Africa, and Europe. For the first time, business actors would establish significant linkages, not just to export or to invest individually, but through their entrance into economic, social and political networks within the member countries. This would allow them to attempt a consensus within the networks on the one hand and to broaden the basis for an oligopolistic competition on the other 39 . Turning now to the regional policy-making face of business actors' strategies linked with the new competition, the issue of sectoral accords (SA) comes to the forefront. The cross participation in privatization in both countries was one of the two main symptoms ofa steel regional restructuring. The other was the steel SA, already mentioned in the first section of the paper. Signed in 1992, it was "surely accompanied by another one not formally approved but more concrete" (Lopez et al.,1994:124-5). What was this accord about and why was it the only one successfully approved, while the petrochemical one failed? A steel manager described it. "We considered, between private sectors, and it was agreed: no price controls; no new subsidies; no price discrimination (clarifying that this agreement was postto the steel agreement, and that it means Brazilians would sell to Argentineans at the same price they sell to their Brazilian alliances); and no unfair competition of dumping" (FWI 61: 19). The main objective behind it was to allow the sectoral networks to be worked out, as concretely affirmed by Acindar's president in a paper (Leone, 1994) From within the steel sector, explanations were that industrial policy restructuring was an issue "left to the firms themselves within the SA ... But the SA were not applied directly" (FWI 59:26). For a steel manager from Argentina, it was in reality a period of cooperation "technically called SA" (FWI 61:2). "The issue was how to generate a framework in which a predatory position would not be allowed" (FWI 61: 19). Its premises lay on plant specialization and privatization. A steel manager from Brazil had a similar view about the SA: "firstly, cooperation was sought, but trade issues prevailed due to the conjuncture, the opening to imports" (FWI 97:9).
In 1991 and 1992, together with the brunt of unilateralism, there was a downturn worldwide cycle with oversupply and low prices for steel and petrochemical products. Brazil's steel sector, as explained, was 8.5 times larger than Argentina's, which perceived it as a 'death threat'. In the process of regionalism, this was a period of transition from the ABEIP towards the Asuncion Treaty's establishment of automatic liberalization across-the-board. National coordinators of the WG7 on industry in the two countries gave related answers: "It was approved because of the moment. The petrochemical one had demands within economies under full deregulation. Thus, the latter was rejected." (Argentine WG7 coordinator, FWI 4:16). "They [steel] arrived first, and were approved as part of [Mercosur and the ALADI system] as a bilateral agreement" (Brazilian WG7 coordinator, FWI 13:12). Timing and being the prime mover were essential. From within the Argentine ministry of Economy, "it was proposed to them [business] in the beginning, later ... [i]t was dangerous in a moment of opening and privatization. It would have decreased the degrees of freedom of movement. If there were necessary changes at the macroeconomic level, the public sector would have been tied by hands and feet" (FWI 62: 11).
Academic researchers explained the SA as policy instruments chosen by the governments and offered as a "double-way strategy for [regional] negotiations" (FWI 32:21,also Dr. Lavagna,FWI 35:18). The acceptance of the steel accord was eased because it seemed to follow the EU steps. "The sectoral accords start with sectors with a high concentration and scales of production coefficients. The most classic case was the EU with CECA" (ibid. 35:10). Moreover, ECLAC's proposal for "open regionalism" stated that one key mechanism was the sectoral accord, "the reduction of transaction costs and discrimination within the region could be reinforced by sectoral arrangements or policies to take advantage, in tum, of the synergetic effects of integration" (ECLAC, 1994:13).
SA have been described as "an instrument of managed trade for reciprocal protection" (FWI 32:20) and as "fragmented business initiatives with specific interests and a vague regulatory framework" (FWI 32:24). Even if this is correct, there seems to be a more fundamental underlining strategy. The SA worked in fact as an open space for the private sector to handle the existing asymmetries (structural and regulatory) by themselves, which the states were unwilling to undertake and for which states invited business actors' involvement. By so doing, the issue of the SA encompassed the process of privatization and restructuring of the first half of the 1990s, allowing at the same time business to build regional networks.
The petrochemical SA was rejected, mainly because of obstacles in tern1S of contents and timing. One of the objectives of the petrochemical sector was to defend the market shore against extra-Mercosur imports (Argentine petrochemical consultant, FWI 47:23). They wanted an automatic implementation of anti-dumping measures, and this was not compatible with the Asuncion Treaty. At that time, international competition was stronger due to DCs recession, low demand and lower prices, like for steel. After the Cavallo plan was implemented, though, the issue of international reputation was at stake, given the government's international commitments to unilateralism. The government began to regulate the issue of anti-dumping only in 1993. Finally, the agreement was rejected, while the private sectors stopped lobbying for it since there was little time left for 1994, when Mercosur would enter a new phase as a customs union. "SA were just a means to regulate the transition to a common market. Argentina's government answered that it was fine in 1994, when there was only one year left. Brazil's government never gave a response. The SA project had been proposed in mid 1993, I believe. So what happened is that they [the sectoral actors] desisted and did their work. There was someone in the Secretary of Foreign Trade who did not like the accord. The main real issue [at the time, became] the harmonization of the CET" (Mr. Mariani, owner of a Brazilian holding, FWI 39: 13) Business actors in petrochemicals concentrated thereafter in preparing the proposals for harmonizing the sectors' tariffs and the sectors' technical standards.

The Perception of Mercosur as One Domestic Market: Search for Regular Patterns
Trade liberalization has augmented the potential for constant flows under a uniform and expectedly reliable treatment for the same products in the member countries. Moreover, the establishment of long-term sectoral networks assures constant flows of exchanges, making up for the lack of "hard" vertical integration, decreasing transaction costs and providing for a long term planning. A search for regularized and coordinated patterns of behavior has since evolved, leading to a situation in which national firms and holdings perceive Mercosur as one domestic market.
In 1996, commodities and semi-processed goods represented 43 percent and manufactures 56 percent of total intra-regional exports (Markwald et aI., 1999:65). Petrochemical trade between Argentina and Brazil has increased strongly (see Erber 1995, Table 10, andAbiquim, 1995). A similar trend may be observed in the steel sector, with the difference that Brazil is a net exporter of steel 40 . Intra-industry flows have also increased in the 1990s, appearing consolidated in the chemical and transport sectors, which include petrochemical and steel products. With a coefficient above 40, such intra-industry flows responded approximately for 30 percent of the 1996 total bilateral trade (Machado et aI., 1997:201-204). In 1999, the same researchers have validated the hypothesis that intra-industry growth in the region is only due to Mercosur (Markwald et aI., 1999).
Considering the increase in exports, the high intra-industry trade, and the significant alliances, business actors have not only risen their stakes through regional trade and investment, but they have also sought to ease flows within their cross-boundary networks through harmonization of norms. In 1994, from a total of "224 non-tariff restrictions and measures on imports and 51 on exports" identified by the member countries and set for ham10nization, 24 percent corresponded to technical standards and 40 percent to agricultural policy (Bouzas, 1997: 20). Therefore, many remaining obstacles have been related to non-tariff trade barriers, apart from the structural asymmetries between the two countries in terms of scales of production. The solution to the new needs, in particular harmonization, cannot be solved within the economic sphere, since they have to cover the whole regional space, not only the networks. They must be solved within the public policy making sphere. From the start of "open regionalism", Argentina requested mutual recognition and a later harmonization of norms, but Brazil rejected this option (Argentine WG7 coordinator, FWI 4:4; Argentine WG3 coordinator, FWI 10:16). A de facto arrangement resulted in the enforcement of the national rules at final destination, which, according to interviewees, meant the norms of the largest market -Brazil. How did business actors react to this situation?
In June 1992, as seen in the first section, ten Working Groups (WG) were established to harmonize norms within the region. The WG closely related to petrochemicals and steel were the WG number 3 on technical harmonization, number 7 on industry, and number 1 on tariffs. Coordinators of the WG maintained in the interviews that politicians decided over the technical part, the idea being that there should be no regulation at the supranational level (Argentine embassy in Brazil, FWI 6: 15). The WG3 work was divided into two types to harmonization: obligatory and voluntary (ISO quality and productive norms). Norms could be drafted in two ways, as mentioned in section 1: through consultation with business actors or through business' direct voluntary action: "We worked by consensus and then norms had to be approved by the CMG" (FWI 15 :23). The 1990s policy style in the region was evolving towards cooperation between state and business (Peres Nunez, 1996), and the process of "open regionalism" appeared rather like emerging regional patterns of regulation, for which policy networks worked within a fi·amework. In the very beginning, Mercosur govemments had a top-down and reticent approach to business' full participation, signaling that the process would be automatic. Business participation, however, was gradually recognized as a necessary pan in the new regionalism. On the business actors side, in the beginning, "even though being little [business actors participation], it was enough to capture the functioning of intergovernmental negotiations ... " (Mariani, owner of a Brazilian holding and representative of Brazil's Chemical Chamber, FWI 39:34). In the beginning, the strategic stand of the firms in many sectors was to observe and discover the rules of the game. On the governmental side, it was, above all, a question of information and efficiency in policy-making. Business actors had the technical expertise coupled with inside knowledge on investments and trade in the sector. They had vital information concerning regional harmonization of norms. In order to understand this phenomenon, we need say that the national process of structural reform not only involved privatization of firms, but also a shift towards the private sector of human resources and flows of information, necessary to make policy decisions. As a result, much of the knowledge was no longer in governments' hands. "At the level of the private sector, [the proposal] is to advance in the strengthening of business chambers as intermediate mechanisms, capable of participating in the policy implementation" (Peres Nufiez, 1996: 26). Dr. Lavagna confirmed this process: "The Chambers joined in, the firms helped in the question of information" (FWI 35:28). As Mariani explained: "This was up to one moment, I believe in 1993, when they started the discussion on the harmonization of the customs' tariffs nomenclature. The government recognized they could not work out the chemical sector, because it was so long and complicated. Extra-officially the two governments asked both national sectors ... This never became an official invitation ... Negotiations may have started in 1992. In Brazil it was in the Ministry [of Industry, Trade and Tourism], in Argentina in the Secretariat ofIndustry" (FWI 39: 17).
After such recognition, business actors participated all along the process, through two main types of representatives who conveyed sectoral interests: owners and top managers of holdings and national trade, and industrial sectoral chambers (the late director of the Mercosur Department of the Argentine Industrial Union, FWI 52:22). Business actors participated in the WG 3 and actively assisted WG I, while only the petrochemical sector did so in WG7. An increasing mix between public and private became normal. The WG 7 on industry was perceived as disappointing, since its purpose did not include the drafting of regulatory norms (WG7 coordinator, FWI 4). WG 3 was perceived as successful. Harmonization was discussed and endorsed by the firms and sectoral chambers. Within the WG 3, they could decide the agenda and approve decisions by consensus. The WG 1, working on the common external tariff, was also perceived as successful because the proposals for the harmonization of the sectoral tariff nomenclature and range were drafted by the sectors themselves, through the joint work of both national chambers. The universe of tariffs for petrochemicals was more complex than the one for steel 41 • Cavallo's proposal was to set the petrochemical sector apart and leave it as a free trade sector, perhaps because there was a tacit preference for verticalization with the oil sector. In any case, governments did not know how to maneuver among the multiple conflicting interests without provoking significant political pressures that could halt the rapid evolution of regionalism as an "open bloc". As a result, the chambers engaged in harmonization during two years. "The final report was presented in the first semester of 1994, very dynamic and well organized, after several thousand man-working hours, and was considered a joint project. They used to get together every month. The governments approved 95% of it. However, lately, the governments distorted it with exception lists in the second semester of 1994, in detriment of the logic of the tariffs" (FWI 39: 18). On the CET itself, they proposed a range of tariffs, i.e. the acceptable top and bottom limits of the norm being discussed by the governments. Governments then approved them, with one or two exceptions. As recognized by one negotiator, "the final exceptions respond to the lobby capacity" (FWI 88:13). However, governmental acceptance of businessdrafted norms did not entail state compliance, especially if other priorities (e.g. inflation) were considered under threat.
Putnam's two-levels of negotiations (1988) do not seem directly applicable in this case, because it appears as an inverted double stage. Decisions by consensus within "open regionalism" meant that the first round of negotiations would take place, mainly, among the private business actors. The ratifying stage corresponded to state officials. The national executive branch in both countries did set the framework, but issues would be drafted by the private sector and then presented to the other level -the State. State officials were the ones to debate the draft and to approve or reject it. In the beginning and at the end, states decided. In the meantime, business actors were furthering 'deep integration' (Lawrence, 1996). Significantly, the sectoral chambers' managers recognized that governments still had the 'upper hand' in the process of regionalism. The two main reasons were that, first, governments derive their legitimacy and authority from state sovereignty and from democratic selection processes. Second, government officials enjoy a broader autonomy given the multiple negotiations engaged in the regional policy-making process. O'Donnell's description of the 1990s "delegative" democracies in the region as liberal but less accountable (O'Donnell, 1994) corresponds to some extent to the experience of "open regionalism", as the latter was partly a consequence of the former.

The Search for Certainty: Compliance and Conflict Solution
At the end of 1994, another transition began, in which new demands for regular patterns of interaction sprung concerning conflict solution and effective state compliance. Conflicts mentioned by interviewees when entering the customs union implementation period pertained to sectors' claims against states for not observing governments' public commitments. The reaction of the negotiator for Brazil's chemical chamber to the lack of state compliance in early 1995 Gustified on the basis of anti-inflationary measures) was the following. "The ABIQUIM position was made lmown through our decision and preference for a tariff structure similar to that of Europe and the USA. And not to respect it in the most important products, which have no bearing against inflation! Because the price is now international, it does not create a reliable image. And now, they continue to negotiate" (FWI 39:22). In the 1990s, business has slowly been shifting to supranational mechanisms of private governance, in the search of mechanisms for conflict-solution. Dagoberto Lima Godoy, Co-ordinator of the Brazilian Section of the Mercosur Industrial Council -MIC (a newly created business organization)-explained that, concerning trade conflicts: "the Brazilian section in the MIC proposed to institute a System of Previous Consultation between the private agents and the MIC industrial entities. The idea is to create a private channel, searching for a solution before reaching the governments,,42. This is consistent with 1995 interviewees' responses. Justifications were not only the rate of speed to negotiate a solution and to implement it. Business actors consciously avoided politicizing conflict, which, they believed, would be harder to solve, as the director of the Latin American Steel Institute argued: "In ILAF A they have participated in the dumping issue. These were meetings to prevent reaching the Governmental Way, and to de-politicize the issue" (FWI 59:8). As said in the first section, Mercosur has been gradually accepting a system of private arbitrage and mediation. The pattern of privatization of conflict solution, exemplified in the above excerpts, matches these more recent governmental trends.

v. WHAT LED BUSINESS ACTORS TO BECOME INVOLVED IN MERCOSUR'S OPEN REGIONALISM AND WHICH ROLE HAVE THEY PLAYED?
Mercosur's regionalism has been characterized by a high trade growth together with minimal institutions and regulation. The priority has been the achievement and maintenance of economic stability. In case of conflicts, in "general, solutions were found at the presidential level. The most important factor in Mercosur's transition period has been the political will of Argentina and Brazil to carry on with the regional integration process. Such will had a central role, given the lack of co-ordination in the macro-and micro-economic policies between the two countries, and the choice for a low degree of institutionalization"(Motta Veiga, 1995). This may have been true in general. The picture, though, appears qualified if taking a sectoral approach to the analysis.
The values framing the conceptualization of open regionalism called for the active participation of business actors in the process, as well as the new policy-making style. Business actors, mainly owners and managers of holdings, and representatives of sectoral chambers, represented the sectors in negotiations on regional norms. Their role had several functions: aggregating, representing, channeling intensity of preferences, influencing policy-making, and beyond, when states did not have the information and knowledge most needed to conduct and carry out the negotiations, such as on harmonization of norms. The sectors' representatives, as sociopolitical organizations, appeared to have a division of tasks, supporting the regulatory process. National chambers worked on the CET and harmonization, Latin American ones provided neutral spaces for network building, and the bilateral one worked on the initial drafting and framing of the whole regional integration process (both ABEIP and Mercosur). Business actors have been involved in the construction of the regional market, in the sense that a market is a historical sociopolitical construction, "an institution that should be created through the definition of collective rules that foster positive dynamics among these individual actors" (Bianchi et aI., 1994: 29). In the same sense, the firm can be understood as ~'moment and segment of a population of actors with' a contextual economic goal, who are contemporarily active in the definition of both themselves and the market that is a market of and among organizations" (Sapelli, 1993:120). The firm is then an association among subjects who carry strategies of behavior and of cultures, while its market is the result not only of firms' actions but also of sociopolitical associations with which the firm constantly interacts.
Thus, the political sustainability of the process has relied partly on a complementary process of coordination undertaken by business within the framework of regionalism. Business actors have developed a regional vision of their sectors, supported a national consensus based on the need for a stable macro-economic framework, and been concerned with the lack of policy harmonization. It is now recognized that "the process of trade liberalization requires increased international regulation. To avoid possible conflict between the creation of trade blocs and trade liberalization on a global level, a WTO-related debate has emerged on harmonizing norn1S, standards and tariff regimes" (IRELA, 1997:29). These three issues have been the ones that business actors have focused and worked on. Consequently, pressures to deepen regionalism stem from below. For example, in November 1997, when there was the threat of a financial crisis stemming from the Asian one, the two main industrial umbrella organizations of Argentina and Brazil issued a document called "Sao Paulo Consensus", calling for deeper regionalism and further institutionalization of regional mechanisms, among which the improvement of intergovernmental consultation and a common position in international negotiations (La Ley, PET/Mercosur, January 15, 1998: 20).
Business actors have engaged in open regionalism in order to respond to the challenges stemming from the environment encompassing regionalism. Enlarging economies of scale to the regional level lends support to regionalism as Milner said, and may in no way be contradictory to claiming protection within the national market, especially during a downturn economic cycle. However, there appears to be a more fundamental consideration in the business actors' rationale, based on the 1995 interviews. Business actors rationalized the new type of competition through the concept of "business scales", which envelops and transcends the concept of economies of scales until now applied at the individual plant, production level. This concept was argued in Oliveira's doctoral thesis on the petrochemical sector in Brazil, considering it as a new form of pre-emptive entry barrier, while strengthening the firm's competitiveness as locus of capital accumulation (Oliveira, 1994). The tangible and intangible assets of a holding or conglomerate build the 'business scales'. Assets include capital investment, R&D capacity and technology, autonomy in organizational and managerial practices, trade links and strategic alliances. Business' objectives focus on coordination and enlargement of business alliances and market shares, taking into account the world industry cycles. The justification for the developments in the petrochemical industry in the region has been that "the 1990s competition is inducing the firms to adopt strategies of concentration and of strategic alliances ... due to newcomers from Asia, chemical firms, and the big world consumers of resins like the automobile industry. At risk is the permanence of the current firms, that alone would not survive". Lucianano Coutinho' sectoral study argued that only a petrochemical firm "capable of USD 3 billion per year of sales can compete ". Ode brecht, the largest in Brazil, had sales for only USD 1.2 billion in 199i 3 • If the 1990s regionalism is new in itself, it is so because its "axis in not on trade", as economist Nochteff clarified in a brief discussion with the author in 1995.
In ECLAC's terms, Mercosur's "open regionalism" has been largely successful in reconciling regionalism with trade liberalization. Amidst a world economic restructuring, this meant an insertion into a new type of competition. Regarding the participation of business actors amidst a general democratic deficit in Mercosur, the issue of accountability arises. While the Mercosur states have remained attached to a Westphalian image (Mittelman, 1996: 197), regional networks may provide the necessary information and coordination in policy-making that the former cannot. Business actors' participation has taken place to the extent described in this paper partly because of the lack of mechanisms for representation and participation in policy-making, whether a supra-national bureaucracy or a regional parliament. The question that lies ahead is whether Mercosur's "open regionalism" should limit its "openness" to the economic sphere, or whether extend it to involve the regional population at large. This question bears a systemic implication, for Mercosur's long-term sustainability rests on its legitimacy and the broader support and involvement of society.

Notes
Mercosur covers a total surface area of 12 million square Km (60% of South America). With a population of 200 million people it had a gross domestic product of almost 850 billion dollars in 1995 (Irela, 1996).

2
Alexander Yeats, from the Trade Division of the World Bank, qualified Mercosur as "the most convincing and worrying demonstration until now of the potential adverse effects of RJAs". His work, echoed in The Wall Street Journal, The Journal of Commerce and Financial Times. was criticised for its indicators and postulates (Bouzas,Clarin,Clarin Economico: 16). The World Bank itself, through its vice-president for the Latin Amelican and Caribbean region, Shahid Javed Burld, dismissed it (Financial Times,. A second Yeats' report appeared in January 1997, this time only with question marks on Mercosur. For a reply to Yeats' comments, see Markwald and Machado, 1999. Together they represent 96% of the bloc's GNP (Leipziger, 1997:2), and have ranked first (Argentina) and sixth (Brazil) among the economies with fastest growth rates between 1990 and 1996 (Markwald et aI., 1999: 64) The two countries account for the most dynamic growth rate of Mercosur intra-regional trade. In 1990, they represented 50.3% of trade, while in 1996 they accounted for 67.8% of the total, which in tum had multiplied by four (TIle Economist Intelligence Unit, 1998, Brazil Market Atlas: 84). Regarding petrochemicals in Mercosur, Paraguay has no local production and in Uruguay the supply of petrochemical is velY limited, although there is a chemical industry of some significance which is highly dependent on imp0l1ed inputs (Hasenc1ever and Lopez, 1998). Concerning steel, the situation is similar.
4 These are, for example, the characteristics of the production process, the type of demand, the type of ownership, the speed of technological change and manner of access to technology, the configuration of the industry at national and global levels, and the type of competition among the main actors.

5
Grant and Paterson, focusing on the ISIC 351 of basic industrial chemicals, argue that the chemical industry is the most internationalized in teffi1S of production and trade, and that the sector has attempted to build some international governance system on their own (Grant and Paterson, 1994: Chapter 6). This might be one of the earliest models for recent 'global industries' .
6 Traditionally, steel was not an internationalized sector. Rather, it was a national symbol of state defense and industrial capability, restricted to the national markets. Germany and Japan have led the internationalization of the sector via joint ventures and mergers & acquisitions (Hudson, 1996:159), and via standardization of products and management connected to the automobile sector (Beddows 1990).
7 Namely, the Brazilian Groups of Odebrecht, Ultra, Suzano, Vila Velha, Ipiranga, Mariani, Peixoto de Castro and Economico for petrochemicals, and Usiminas, CVRD and Gerdau, among others for steel; and the Argentinean Groups of Perez Companc, YFF, Astra, for petrochemicals, and Techint and Acindar for steel.
10 Percy Mistry (chairman of the Oxford International Group) stated: "The new regionalism has only become possible in an ethos of unilateral trade liberalisation, and if that ethos had not occurred, then we wouldn't even be talking about the new regionalism" (Mistry, 1995:30-31). See also Tussie (1998: 91).

12
A complex industrial restructuring has taken place at a global level since the mid-1970s (Dicken, 1998:437-438;Ruigrok and Van Tulder, 1995). "The globalisation of the world economy passes through the reorganisation of processes of production along a rather regional, even global basis ... The importance of economies of scale, that imposes the reduction in the number of units of production, and the need to standardise the processes of production ... combine themselves in order to determine the new modes of articulation of products and processes of production" (Fontagne et aI., 1996: 11). Within this phenomenon, the intermediate or middle goods deserve a particular attention. Defined as all goods produced, re-introduced during the productive cycle and disappearing as such within the cycle itself, they comprise, among other sectors, steel and petrochemicals, their raw materials, and their immediate by-products (ibid: 27). The importance of intermediate sectors in globalization resides in that their percentage in world trade is ever growing. In 1996, they accounted for between 50 and 60% of the world trade among Western Europe, the USA and Japan (ibid: backpage).

13
"In Brazil they must make an integrated industry. But with imagination, it can be reached" (F-W I 5:30, interview with Dr. Amilcar Pereira da Silva Filho, chemical engineer, consultant of Ciquine and Polialden in 1995. Dr. Jovelino, another manager of a petrochemical firm in Rio de Janeiro, added that in 'Mercosur will permit those [sectors 1 to adjust better, in a process of integration of capital" (F-W I 60: 12).
14 "Intra-zone [regulation 1 is the fundamental key to the harmonization of the process of formation of capital" (Aldo Ferrer, Argentinean economist, F-WI 57:6).

15
"There is an important point: the political will of governments was not enough, it was necessary to integrate capital" (Ms. Brun, Brazil's alternate coordinator ofWG7 on industry, and representative of the Ministry of Industry; F-W I 13:6). This was the explanation for entering into negotiations on the only regional (bilateral) policy instrument targeting an industry, the automotive one: "The Automobile Accord is being negotiated because Argentina has a regime, while Brazil has not. This situation could not persist, because it would have created a problem in investments. And because Brazil wanted to attract capital investment after the Plan Real, and the shock of the Mexican crisis" (ibid: 13:41). In 1998, "automotive, auto-parts, cargo and transport vehicle trade continued to be one of the major focuses of bilateral trade between Brazil and Argentina. This happened under the protection of special sectoral treatment, as was outlined dUling the first years of Mercosur. In 1997, sectoral flows reached US$ 4.6 billion and accounted for more than 31 % of bilateral trade.
Maintaining the rate of expansion that the automotive sector between Argentina and Brazil reached during the first six months of 1998 would suggest that by year-end it could reach US$ 5 billion. This figure would increase sector involvement to 33-34% of bilateral trade" (Mercosur Report No.4, IDB -INTAL).

16
The secretariat had both its first Director (appointed in a Mercosur meeting) and its first autonomous budget for 1997-1998.

17
'To date, exchange of information and consultation on macroeconomics matters among governments of the member countries have been informal and linked through personal channels rather than bureaucratic institutional routines" (Bouzas, 1997: 6-7).
18 "The great advantage of the Common Market is that sovereignty is in partly abdicated. What is important is the harmonization oflegal and tax issues, to level the playing field" (Candal, main advisor to ABIQUIM, the Brazilian Association of the Chemical Industry, 21 :3).

19
The Argentinean Sub-Secretary for American Economic Integration affirmed that the sectors under negotiations were chosen according to the sensitivity that their production and restructure bring about (Sabra, 1993: 551).

20
While in 1995, Brazil's steel sector was 8.5 times that of Argentina, analysing the three-year period 1994-1996, it is evident that Brazilian petrochemical production was approximately nine times greater than that of Argentina, visible consumption was six times greater, and exports were five time higher. (Hasenclever and Lopez, 1998

23
These measures concentrate on a!locative efficiency and tend to disregard technology and social relations (i.e. with labor). In addition, they may have as much to do with unilateralism and the national stability plans, as with regionalism. Tradables have indeed shown a reduction in the price-cost margins, and have grown much slower than non-tradables, due to the stabilization plans in both countries. Regarding productivity, steel in the region shows high absolute competitiveness (relative to the USA), while the plastics sub-sector in Argentina has improved in TFP (World Bank 1996, in Leipziger et aI., 1997.

24
The CET "has a simple structure, involving applied ad valorem duties which originally ranged from 0 to 20%, with a horizontal average of 10.1 % and an import-weighted average Botafogo, the Brasilia Protocol prevents "costly bureaucratic structures, and can be used by governments or physical or legal persons". The only clarification regarding the three stages is that the third one of arbitration is ad-hoc, obligatory and without appeal.

27
In 1996, the Brazilian national parliament approved a law legitimizing private arbitration. And in the Mercosur presidential summit of July 1998, the ministers of justice accepted a system of international commercial arbitration as the alternative to solve private conflicts regarding international contracts. Mercosur signed two legal instruments on the issue: one among the member countries of Mercosur and another between Mercosur members and associate countries (AFP-Agence France Press, July 20, 1998). Moreover, Argentina and Brazil have recognized the autonomy of the will of contracting business parties in deciding the law that will be applied to their commercial relationship. In 1998, Uruguay was debating whether to implement it, in order to harmonize the regulation within Mercosur ("Parlamento uruguayo estudia convencion sobre contratos internacionales. Uruguay pone al dia su legislacion por el Mercosur", in Sucesos Mercosur news, July 21, 1998).

28
"Liberals in government thought of RI in terms oflocking-in the reforms at a supra-national level" (former Secretary of Economic Policy of the Brazilian Finance Ministry in 1994, Dr. Winston Fritsch: 30: II).

32
A representative ofthe Chamber of Plastics in Argentina said: "The relation with Brazilians started in 1986, when we perceived the globalization process of the economy, and therefore, ourrole in the process has become very active" (15:1,1:7).

33
It was the Bilateral Argentine-Brazilian Trade Chamber (BABTC) that took Menem in his first visit to Brazil. "Then, the first Menem's trip to Brazil was a charter organized by this Chamber, based on the silent work previously done. Only this year came to be publicly known" (42:10). Both the petrochemical and steel sectors had been members of the BABTC for about 20 years, from the time of ALALC\ALADI. Besides, the previous BABTC director to the one in 1995 was from the chemical sector. The involvement of this chamber in regionalism, though, had a longer history. Its manager affirmed: "Sarney was a person with vision, very close to business people. In 1984, the work for the ABEIP began in our Chamber" (42:38).

34
In 1995, Kosacoff, ECLAC's economist, described the new typical stylized firm in the region: a) from an industrial firm it shifts to incorporate assembling plus its own system of commercialization and distribution. This leads to the situation that profit-making in' Mercosur in the short run is higher. This also happens because of the non-tradables [effect].
B) The entrepreneur is more sophisticated, and has simultaneous management and financial representatives abroad, joint ventures on inputs, and market allotment. This rationality responds to the international, macroeconomic and regulatory framework. It leads to the expulsion oflabor and there are no externalities (J!-W 134:37).

35
APLA, is the sector-based organization that appeared 15 years ago to promote the petrochemical business. First, they waited for a strong political signal, e.g.: in terms of creating one common market, to change their business strategies. Then, its structure and role changed to meet the two most pressing needs for the sector: RI and environment. APLA's structure already offered advantages (interview with the manager of APLA). "Its role changed. There is a delicate game. The idea is not to interfere with the sector-based national chambers, but to offer a space for them to meet (a neutral space)". For that it changed its statutes and created CEQPLA (Business committee including the chemical sector). This committee had the goal of establishing a common language and gathering the executives of firms and the chambers of the region.

36
The first venture was signed in September 1995, in order to build a new unit of styrene in Southern Brazil, producing 180 thousand tonnes, and had Brazil's Petroflex as a shareholder. Petroflex produced synthetic rubber. The idea was that Copesul of Odebrecht would provide the raw materials (butadiene for synthetic rubber), while Petroflex and Perez Companc would manage the demand. The second venture dates from 1996, when Copesul and Perez Companc signed in Buenos Aires an agreement to build a production unit of polystyrene in Copesul, of 100 thousand tonnes per year. The new unit, for which they invested US$ 90 million, operates as from 1998.

37
"Petrobras, headquartered in Rio de Janeiro, has lined up dozens of international and local partners as part ofthe new competitive environment in Brazil's oil sector. A most recent deal to develop fields in the Santos and Campos basins in Southeast Brazil was signed with three partners: Odebrecht, U.S. oil company Amerada Hess Corp. and exploration and production company British-Borneo" (Reuters, February 4, 1999).

38
Odebrecht was present in 20 countries and four continents in 1995, in the areas of construction and engineering, chemical and petrochemicals, infrastructure and public services and cellulose. TIle structure covered 125 controlled and allied firms, under a holding, a group of firms in the first three areas and a development undertaking in the last one, a finn in Insurance, and the Odebrecht Foundation. In 1997, Odebrecht had approximately 40 thousand employees, of which 29 thousand in Brazil and the rest abroad.

39
Not all holdings have built long-term networks. Others have had a more individual strategy, opening offices for sales and services in the region. This type of strategy has sometimes been the result of failed attempts in creating long-term alliances, and, at least in one case was also aimed to obtain inputs at lower prices (FWI 45:14,and FWI 85:87). Finally, when there were few plants producing the same product, directly control has been sought at the regional level One case was Argentina's company (Maleic) bought by Brazil (Group Ciquine) through a majority equity venture.

40
Although the importance of each country for the other is low due to Brazil being a world !cader in steel, and the intTa-indusny trade is also small, changes began to take place in 1993. Compared to 1992, Argentina doubled its importance in Brazil's total steel imports to 9,9%, while Brazil accounted for 48,2% of total Argentine steel imports in 1993 against 46,5% in 1992. Thus, the coefficient of intra-industry trade rose from 7 in 1992, to 18 in 1993. Given the extremely high intra-industry trade between Argentina and the rest of the world of 92, a special report on the region's steel sector anticipated the possibility of regional specialization and co-operation (Nofal, 1994).

41
There was more competition between Argentinean and Brazilian firms, and more actors involved within each national sector. This was caused by the different structure of production in the two countries. In addition, due to the privatization process, there was more competition for control and verticalization in the sector itself, concentrated in the expansion of the size of firms and the integration in two ways: with the raw materials sector and among the national petrochemical complexes. Moreover, there were different national regulatory frameworks. The regulation on raw materials was different in each country. YPP was privatized, while Petrobras remained governmental amidst the privatization of Brazilian petrochemical firms. Thus, YPP, for the first time, entered the petrochemical sector, while Petrobras, for the first time, became temporarily disengaged from it. In the 1990s, average tariffs have been higher in Brazil in basics and plastics, although the situation has been reversed when considering the Argentinean statistical tax of 10% (the latter has been merged within Mercosur's CET in 1997).