ARE REGIONAL BLOCS LEADING FROM NATION STATES TO GLOBAL GOVERNANCE? A SKEPTICAL VIEW FROM LATIN AMERICA

I . INTRODUCTION Regionalism was once thought to be an intermediate step towards global governance. Through it, nation states would gradually transfer sovereignty upwards to the regional level . Regional groupings would therefore constitute stepping stones, as opposed to stumbling blocs, in an increasingly regionalized world system of governance . However, the global stage still is mainly configured by both powerful and failing states as main actors, while regions —the European Union included— have hardly met the optimistic expectations raised just a decade ago . This article analyzes the experience at regionalization in Latin America, assesses the level of interdependence and balance of power among the countries of the Western Hemisphere, and illustrates its insights through the scrutiny of one representative case : Mercosur. The main conclusion is that, at least in Latin America but probably also elsewhere, there is a growing gap between rhetoric and accomplishments as regards regional integration.


I. INTRODUCTION
Regionalism was once thought to be an intermediate step towards global governance. Through it, nation states would gradually transfer sovereignty upwards to the regional level . Regional groupings would therefore constitute stepping stones, as opposed to stumbling blocs, in an increasingly regionalized world system of governance . However, the global stage still is mainly configured by both powerful and failing states as main actors, while regions -the European Union included-have hardly met the optimistic expectations raised just a decade ago . This article analyzes the experience at regionalization in Latin America, assesses the level of interdependence and balance of power among the countries of the Western Hemisphere, and illustrates its insights through the scrutiny of one representative case: Mercosur . The main conclusion is that, at least in Latin America but probably also elsewhere, there is a growing gap between rhetoric and accomplishments as regards regional integration.

H. THE HISTORY OF INTEGRATION IN LATIN AMERICA
Worship to the alleged natural unity of Latin America has been a permanent component of the regional political discourse . As Mace (1988: 404) points out, "the wars for independence were not yet over before proposals for political unity began to be heard throughout the newly independent territories." Simón Bolivar, the Venezuelan liberator, established his belief in a sort of United States of Hispanoamérica in his messages to the Congresses of Angostura (1819) and Panama (1826), which can be fairly considered as the first attempts at regional integration in the continent.
As pervasive as the idea of a continental union could appear, the real unity of Hispanic Latin America was never but a myth . Although it was true that most of the region shared a common culture, language and religion, the divisive factors at work long outweighed those pushing toward unity . Among the natural obstacles, large distances were paramount -be them by ground or by sea. Geography made communications extremely difficult, and contemporary technology was not enough to overcome such a shortcoming . Still, the social obstacles were even more serious : Spain had developed an administrative system aimed at extracting resources and controlling the territory from a single center, that of Spain itself. Consequently, its American colonies were seldom connected to one another, and territorial as well as regulative disputes were conducive to jealousy, rivalry and competition between them . This is why the end of the independence wars frequently led to civil strife, and conflict over borders has become an enduring source of problems that some Latin American countries still face today.
Administrative necessities, along with the impossibility of arriving at any kind of confederate arrangement, ended up in the division of Hispanic South America into nine independent countries out of the originally established Viceroyalty of Peru. Central America, though of a much smaller size, followed a similar pattern of fragmentation, Mexico standing as the only original viceroyalty that managed to conserve most of its territory. On the other hand, Portugal's larger colony -Brazil-kept its unity in spite of its huge extension and its many internal differences . In part, this was due to the fact that the Portuguese Imperial Court was directly transferred to Brazil between 1808 and 1821, thus contributing to the centralization of power -and to the legitimization of the hence strengthened central government.
The numerous obstacles to cultural and economic interrelations faced by the Latin American countries, combined with the failure of the attempts at political unification throughout the nineteenth century, led to a decline in support for the idea of integration. It was later replaced by pan-Americanism, a water-downed concept of continental unity for the management of international relations in the region, from the 1870s to the late 1950s. After World War II, the emphasis was dramatically changed from political unity to economic integration . Many factors were responsible for this turn, among them the functionalist argument that international organization -aimed at world peace as ultimate goal-would be better served by functional arrangements in economic, social, and cultural affairs than by political or federal integration . Such reasoning notwithstanding, the decisive thrust toward Latin American integration came from the newly instituted United Nations Economic Commission for Latin America (ECLA, or Comisión Económica para América Latina -CEPAL) (Wionczek, 1970 ;Mace, 1988).
ECLA's proposal was aimed at the enlargement of national markets through the constitution of a Latin American common market. The coalition of technocrats and reformist politicians led by Raúl Prebisch, then the Commission's executive secretary, considered that economic cooperation was the only means to overcoming traditional dependence on primary commodity export trade. The rationale behind such project was the understanding that the heretofore model of development -so-called importsubstitution industrialization (ISI) -was reaching its limit of exhaustion within the national markets . Therefore, larger markets entailing economic diversification and technological modernization were an urgent need in order to advance further development across the region (Wionczek, 1970). A second, rather political goal sought by the proponents of integration, was to build regional blocs able to counterbalance US dominance in the region (Grugel, 1996). Other accounts also mention the creation of the European Community as triggering the integrative efforts, on the ground that the resultant trade diversion having place in Europe was damaging Latin American countries that were primary-commodity exporters (Mattli, 1999).
The drive for regional integration brought about two waves of integration efforts in the following decades . The first one saw the establishment of the broad Latin American Free Trade Association (LAFTA) and the narrower Central American Common Market (CACM) in 1960; the second one led to the creation of the Andean Pact (later CAN) in 1969 and the Caribbean Community (CARICOM) in 1973 (Mace, 1988). Although some of these processes reached an apparent success during their early years, and most of them are still in existence -whether slightly or radically changed-, none achieved the initial objectives . By the late 1980s Latin America still was an underdeveloped and highly unequal region, and its national economies had neither grown enough to improve most of its population's living standards nor become complementarily specialized as planned. Moreover, the dependence on, and unbalance to the United States had not withered away but deepened further. Attempts at integration in Latin America failed due to different factors. Most significant among them were a low level of prior interdependence, scarce convergence regarding levels of development and economic policies, political instability (reflected in frequent democratic breakdowns), and the adoption of either too weak or too rigid regional institutions. Table 1 classifies the four factors of failure cited above according to two criteria: character (either political or economic) and scope (either domestic or regional) . The historical mismatch between economic interests and political arrangements has frequently been noted, especially along the last two decades. Grugel and Medeiros (1999 :56) argue that "integration in LAC [Latin America and the Caribbean] failed in the past because of a lack of real economic interests between LAC firms, despite the complex de jure structures of integration that were created." Likewise, Mattli focuses on demand and supply of integration as sine-qua-non conditions for success, the latter being indisputable political. He argues that, once "political leaders are willing to initiate an integration process, chances of sustained success are greatest if two strong integration conditions are satisfied : first, a regional group stands to reap important gains from integration ; second, the group is led by a country able to serve as an institutional focal point and regional pay-master" (Mattli 1999 :64-5). By 1990, these conditions seemed to be finally in place: the Andean Community and the Central American Common Market were re-launched and Mercosur (Common Market of the South) was founded.
In 1997, the Andean countries signed the Protocol of Quito in order to flexibilize the bloc internal procedures. The renewed institutional structure was set to be larger and deeper than any other in the hemisphere. In 1990, the institutionalization of the Andean Presidential Council as top decision-making body consolidated the ante . A similar path was followed in Central America : after the end of the civil wars that had ravaged several countries in the region, the democratization processes fostered new attempts at integration. The Protocol of Tegucigalpa, signed in 1991, reconfigured the institutional structure and established presidential summits as utmost power site. The same year, the Treaty of Asunción created Mercosur. Learning from the experience of the neighboring processes, the signatory countries opted for minimizing bureaucratic apparatuses and routinizing presidential summits since the very beginning . In all three cases, demand conditions for integration were on the rise given the new democratic stability and the ongoing processes of economic liberalization; in turn, supply conditions were provided by the dynamics of presidential summitry rather than by supranational institutions . The progress of Latin American regionalism -albeit little-took root in the 1990s . It was due to political rather than economic factors (Malamud, 2004) . But could politics be enough to fuel regional integration and sustain it overtime? III. PROSPECTS FOR ECONOMIC INTEGRATION IN LATIN AMERICA Formal trade arrangements (free trade areas, customs unions, etc) are neither a necessary nor a sufficient condition for fostering interdependence: although removing barriers to trade is ceteris paribus bound to have a positive effect on trade, this might be offset by other factors generating even more trade in different directions (complementarities, economies of scale and scope, comparative advantages, creation of infrastructure, changes in tastes, demographics, income and so on) . For instance the Mercosur countries experienced an increase in interdependence in the years prior to 1999, and a decrease or stagnation of interdependence since then. At the same time their interdependence with China has grown : this country represented about 2% of the external trade of Mercosur by 1993, a proportion that has grown up to 7% during 2005, in spite of the absence of commercial arrangements between China and the Mercosur countries.
Basically, Latin American countries export similar products that are demanded mostly elsewhere in the world as raw materials : two-thirds of all South and Central American exports are either agricultural or fuel and mining products. Although host to roughly 7% of the world population and 5.4% of the world economy, this region only accounts for 3 .1% of the world exports and 2.9% of the world imports (WTO, 2005). Therefore, the prospects for further regional integration are very limited by a relatively small market size and a relatively narrow export base.
Trade interdependence with the United States and other NAFTA countries is high for several South American countries, especially the CAN-5 ones. In spite of all their flaming rhetoric against the US and the Free Trade Agreement of the Americas (FTAA) process, both Bolivia and Venezuela rely on hemispheric trade rather than intra-bloc commerce . This is also the case of Peru, Ecuador and Colombia . NAFTA countries are also important for Brazil and Argentina as trade partners, and even more for the smaller Mercosur members . In contrast, the excellent economic performance of Chile under social-democratic governments has contributed to the belief among several Latin American governments that pragmatism rather than ideological premises should be the starting point when negotiating trade arrangements. The rationale for FTAA for Latin American countries was to increase market access for their producers to the richer North American consumers. After the collapse of the FTAA negotiations, Colombia and Peru have ratified free trade agreements with the US, following the path of Chile and the Central American countries . Ecuador and Uruguay may follow suit in the coming years, and something similar could happen in Paraguay and even Uruguay, where the idea is being publicly discussed . Greater interdependence with the rest of the world is also likely to come from new trade agreements with countries outside the western hemisphere, particularly in Asia. Chile has already signed treaties with South Korea, China, and other Asian countries, while some governments in the region have expressed their interest in doing alike . As the Uruguayan government has seen, there is a political advantage in proposing trade agreements with Asian countries: when seeking many different partners appears to be the goal, positions based on ideological grounds (which comprise a significant part of the opposition to LAC-US, or even LAC-EU trade agreements) are likely to become weaker . The development of more and more bilateral arrangements with the rest of the world, and with NAFTA or Asian countries in particular, is expected to lead to greater trade interdependence with these regions, and therefore, to weaken the prospects of greater regional interdependence.
As shown in Figure 1, Mercosur is a relatively closed economy, although it is becoming less so . Still, trade as a share of the economy is very small . The period 1991-98 may in the future be called the "Golden Age" of Mercosur . Total trade of member countries grew from 7% of GDP to represent more than 11% . Mercosur underpinned this increasing openness of its members' economies: trade between Mercosur-4 and the rest of the world was about 8 times that between the bloc's members in 1990. By 1998, this ratio had dropped to 3 .3. With the advent of economic crisis, first in Brazil, and later in Argentina and Uruguay, both openness and interdependence fell between 1998. Between 2002, an unusually benign international environment -with record cheap credit for emerging markets and record high prices for the commodities they export-has enabled Mercosur economies to grow again . Openness has increased and reached new heights, but this time it is the rest of the world what underpins this trend. As a result, trade with the rest of the world grew to be 5.5 times intra-bloc trade during 2005. As pointed above, after more than a decade the tendency for interdependence is far from being consolidated towards an ever closer union, at least in the commercial realm . What does the picture at country level look like? For Uruguay, trade interdependence with the Mercosur-4 bloc peaked in 1998 and then fell consistently . No wonder then, that the country is considering signing agreements with the US and other countries

12%
Economic Crisis Bra I Ag 1 Unfailing interdependence and openness ---------------- -1990 1991 1992 1993 1994 1995 1998 1997 1998 1999 2000 2001 2002 2003 2004 2005 outside the region. In the case of Brazil, trade with Mercosur countries represents today a similar proportion as before the founding treaty was signed. Mercosur has indeed become a more important trade partner for Paraguay and Argentina since the inception of the bloc, but as shown in Figure 2 the tendency seems far from consolidated . In fact, Paraguay may now be following the pattern Uruguay has exhibited since 1998. Officials and analysts of Mercosur frequently liken it to the process of European integration. They do so for myriad reasons: some explain the development of Mercosur as a form of institutional mimetism (Medeiros, 2000); others take the EU as a yardstick against which similar processes are measured (Camargo, 1999) ; and others attempt to draw lessons from the EU about what can or cannot be done (Flôres, 2004 ;Malamud and Schmitter, 2006). It is now clear that institutional mimetism has not taken place, and that the EU is not a standard model but a unique phenomenon that can serve as a point of reference and inspiration -but that cannot be transplanted or replicated wholesale . Indeed, most EU-worshippers in Latin America usually fail to distinguish between process and result . In the EU, transnational transactors and national governments demanded institutions and rules, which were supplied by national governments, the European Commission, and the European Court . It is impractical to understand the emergence and development of regional institutions without reference to demand and supply conditions . However, political debates about Mercosur have become more and more ideologically loaded, which contributes not only to obscure comprehension but also to harm the cause allegedly defended.

0%
Two discourses on Mercosur can be distinguished in the literature and public debates . One, which could be labeled as progressive, existentialist or developmentalist, is best incarnated by such intellectuals as Hélio Jaguaribe (2001) and Aldo Ferrer (2001), and top officials of the Lula administration are said to sympathize with it . The other, best nominated as realist, instrumentalist or liberal, is expressed by scholars such as Paulo Roberto de Almeida (2003), Roberto Bouzas (2002) and Bouzas, Pedro da Motta Veiga and Ramón Torrent (2002) and found closer affinity in the previous Brazilian administration headed by Fernando Henrique Cardoso. The differences between these discourses can be illuminated through a comparison of four key dimensions.
The first dimension regards the underpinning of regional integration. While the "progressives " claim that regional building should be based upon ideology and high politics -mostly related to power balancing and global order-the "realists" focus on interdependence and low politics -mostly related to factors such as trade, investment and macroeconomic coordination.
The second dimension concerns the cement of integration . What is it that holds a bloc together? The progressives sustain that a common identity constitutes the indispensable glue that keep different societies in concert towards an ever closer union . In contrast, the realists argue that a process of voluntary integration among sovereign nations can only survive if it is held together by concrete interests, be them national or, especially, sectoral.
The third dimension is the diagnosis of the current state of integration in the Southern Cone . Following from the above positions, progressives argue that Mercosur suffers from an excessive emphasis on commercial aspects whereas its leaders pay very little attention to the, allegedly more relevant, social and political aspects. Realists, in turn, state that Mercosur main deficit is not the democratic one ; instead, its main troubles involve defective economic integration and an excessive degree of politicization, understood as the political management of technical issues.
The fourth dimension deals with the proposals for overcoming the current impasse . The progressives defend two main priorities : an increasing participation of the civil society and the establishment of representative institutions that foster regional democracy . The realists call for the strengthening of technical proceedings and higher functional integration, especially regarding physical infrastructure .
The above comparisons have been stylized for the sake of clarity. Sometimes, however, the differences between the groups are more nuanced . This is the case as regards regionalizing energy and transportation networks, which both promote -albeit with different emphasis . There are also some analysts with more tempered visions who, such as Félix Pena (2006), are moderately critical while advocating operative, middle-term proposals for overcoming the current pitfalls, but they are very few. Bridging the gap between the two groups is not quite an attractive job.
There is still a third group, mainly composed of businessmen and neoliberal economists, who campaign for the dissolution of Mercosur altogether -or at least for abandoning the goal of the customs union and keeping the bloc limited to a free trade area . The rationale behind this thought is that economic integration, understood almost exclusively in economic terms, would benefit the Latin American countries as long as the most advanced economy in the hemisphere were a part of it . For this view, Mercosur is nothing else than an obstacle, not a building bloc, to a free trade agreement of the Americas . Domingo Cavallo in Argentina and some spokespersons of the powerful Federation of Industrialists of São Paulo (FIESP) backed this position. This paper does not deal with it further because such a view does not shed any light on the phenomenon herein studied, except for claiming its perversity.
The remaining of the section addresses two set of proposals . First, it discusses some "progressive" ideas that promote a deeper politicization of Mercosur and the establishment of supranational, representative institutions. Next, it introduces some suggestions that have been advanced by "realist" or, it could be said, neofunctional ' analysts of regional integration.
Progressive admirers of the European Union champion the creation of supranational institutions in Mercosur such as a parliament, a tribunal and a commission . These are fascinating proposals but very unlikely to be implemented in the short run.
The establishment of a supranational parliament is expected to confer democratic legitimacy on a Mercosur that supposedly suffers from a democratic deficit. Mercosur decision making, the argument goes, is opaque, distant from the people and more concerned with economic matters than with social and political issues . The projected solution is to promote democratic legitimacy with the direct election of representatives. A regional parliament, it is claimed, would foster popular participation and citizen representation. However, it is a daunting feat to design a feasible composition of this body and explaining how it could preserve a balance among the member states and, more importantly, what decision-making powers it would actually have . Three issues need to be addressed to determine the need for and viability of this proposal : first, is there really a democratic deficit? Second, how could a parliament lessen that deficit? Third, what composition and competences would a Mercosur parliament have? Regarding the first point, it is hard to find a democratic deficit in Mercosur: since its decision-making process is entirely intergovernmental all decisions are made by consensus among officials of the four national governments, all of which are democratically elected. Moreover, decisions are not valid until approved by every member country following domestically determined procedures. So Mercosur is as democratic as the most democratic of its members . As to the second point, it can be argued that a regional parliament may anyway strengthen transparency by fostering public debate and helping to bridge the gap between the authorities and the citizenry . If so, the question is whether a parliament can make this happen. According to the available data, public opinion in all Mercosur countries locates parliaments, parties and politicians at the bottom of a scale of credibility (Lagos, 2002) . It is hard to imagine how these unpopular institutions might foster credibility and legitimacy at a regional level . The third point raises the issue of viability rather than need: first, as regards composition, sharp population asymmetries make this a difficult puzzle to solve . Brazil has roughly 80% of the bloc's population, so any distribution that allocates to it less than 50% of seats will be perceived as undemocratically biased and will, at any rate, face resistance from Brazilians motivated by considerations of national interest . And yet, giving Brazil more than 50% of seats means that one country would permanently hold a majority against all the others. A compromise could be reached by conceding either party the majority, and by simultaneously stripping the majority of any real power, either by requiring super majorities or by not endowing the parliament with significant competences. Whereas the former option diminishes democratic legitimacy and popular sovereignty, the latter deprives the parliament of effective decision-making capacities. Neither option is a solution therefore.
The aforementioned does not preclude the establishment of a parliament sometime in the future, but this would have to be preceded by the enlargement of Mercosur and a mitigation of current demographic asymmetries. Failing this, the most likely outcome is a parliament that will be toothless, a deliberative forum rather than a decision-making body. Given this context, it is hardly surprising that there have been few demands for the parliamentarization of Mercosur from any significant social or transnational actors. To be sure, previous Latin American experiences at parliamentarizing regional blocs are far from promising . The Andean Community and the Central American Integration System have both established supranational parliaments, but comparative studies show that their performance have been disappointing to the point of irrelevance (Malamud and Sousa, 2005).
As for the tribunal, Mercosur has had an ad-hoc dispute-settlement mechanism since its inception. It provides for a three-stage process of direct negotiations, mediation by the GMC (Common Market Group), and arbitration by an ad-hoc tribunal . The objective of the institution-builders was to create arenas to settle disputes through juridical rather than diplomatic means, thus creating a third party in addition to, and independent of, the parties in conflict. A main limitation is that, unlike in the EU, only states can be parties to a dispute, so that the mechanism is more like those found in the World Trade Organization than in the EU . The net result is appalling: not only citizens have been denied access to regional courts, but judicial activity -and therefore judicial activism such as that which fostered European integration in the dark ages-has been negligible. Only rarely have Mercosur member states resorted to the dispute-settlement mechanism: whereas the European Court has produced hundreds of rulings every year always since its creation, the Mercosur dispute-settlement mechanism has been used only 10 times in 15 years . Although there appears to be no official or social demand for an empowered court or for the judicialization of regional procedures, a permanent appeals court was set up in 2002 and started functioning in 2005 . However, it is not entitled to adjudicate in a conflict in which a part opts for appealing to WTO proceedings: it is a standing institution whose services are merely optional and have never been asked for. To make prospects cloudier, even if these antecedents were neglected and a supreme tribunal was finally established, the non-existence of Mercosur common law (every piece of legislation must be transposed into national legislation by all four members to be enforced) would render inconsequential any increase in judicial activism.
As regards an executive authority, Mercosur has never had anything other than intergovernmental organs . An Administrative Secretariat was established in Montevideo in 1991 and was turned in 2003 into a Technical Secretariat, but it enjoys neither steering autonomy nor political leverage. There were two main reasons for keeping Mercosur strictly intergovernmental: the fact that national governments did not want to loose control over the process, and the lessons derived from precocious overinstitutionalization in the past, when other regional projects (the Central American Common Market and the Andean Pact) ostensibly failed to deliver (Malamud, 2004) . These reasons remain valid today . Although some heads of state routinely issue public declarations calling for a stronger Mercosur, no national government has signaled a credible intention to relinquish power to a supranational body ; and institutional performance has not improved in Latin America significantly, so institutional inflation as a way to solve problems seems ingenuous at best . A third reason why the creation of a supranational commission is not expectable is that there is no social demand for it: entrepreneurs and transnational transactors are accustomed to appealing to domestic governments, particularly chief executives. Mercosur is in this sense a power-oriented rather than a ruleoriented organization, and as long as vested interests have access to the powers that be, they will see no reason to invest in the creation of new institutions . So, in the absence of social demand and political supply for a supranational commission, it is unlikely that one will be created.
In contrast with the great ambitions that transpire from the above proposals, the realist vision call for more low-key actions oriented at feeding a slow process of gradual integration . With the caveat that not all the prescriptions are shared by every scholar that we consider a realist, these proposals may be grouped into four sets : functional tasks, core areas, technical management and effective implementation . They are discussed next, drawing on Malamud and Schmitter (2006).
The most clearly neo-functional recommendation suggests selecting a functional task or two . The chosen task must be separable, manifestly difficult to realize within the confines of a single national state, and capable of generating concrete benefits for all participants within a relatively short period of time . Two functional tasks are better than one so that trade-offs can be negotiated across them. In Mercosur, the joint energy and transportation infrastructure seems to provide an appropriate and apparently separable set of "functions" . These areas do not require any initial commitment to further integration. They are capable of generating tangible benefits on their own.
A second advice is to select a core area of contiguous units with internal lines of communication and exchange and, if possible, convergent motives for cooperation . In Mercosur, energy, transportation and, more generally, physical integration appear as the obvious areas to invest in. Cross-border infrastructure on roads and ways -on land and water-need urgent upgrading in order to smooth transactions, foster interdependence and develop joint production chains . This requires the establishment of structural funds and could be benefited by a mechanism of opting out such as the one in force in the EU, so as to allow reluctant partners not to join initially while leaving the door open for future inclusion . Border crossing and customs procedures need to be simplified and facilitated, especially for people that live close to the inner frontiers of Mercosur . The difference between being -whether to live, shop, work or study-on one side or the other of a cross-national border should be minimized, and transit facilities are crucial in this respect.
The third dimension regards technical management. Realists state the convenience to start with low-politics and low-visibility projects in order not to attract the opposition of potential veto players . In Latin America the largest one is the US . It might smooth the way seeking to convince the US that these "low grade" efforts will contribute to its broader goal of free trade within the region while not threatening its geopolitical interests . Also powerful domestic actors may be hurt by the integration project and attempt to curb it, so it is advisable to reach a high level of regional interdependence before nationalistic claims are able to mobilize support in favor of domestic closure.
The claim for the establishment of a common foreign and defense policy is, in this view, either unfeasible or counterproductive. As the argument for a common policy goes, Mercosur's goal is to contribute to multilateral governance by becoming a relevant global player, and such a strategic dimension implies the integration of defense and foreign policies. Apart from ignoring the European experience, which shows that foreign and security cooperation is much more complex than low politics cooperation (as exposed by the failure to adopt a common position on Iraq and UN Security Council reform), it neglects the clash between Argentina and Brazil over a key priority of Brazilian foreign policy -winning a permanent seat in a reformed UN Security Council, a claim militantly opposed by Argentina. Closer foreign policy cooperation by Mercosur members is certainly possible, but the top foreign policy issues are likely to remain divisive.
The fourth aspect has to do with effective implementation . The contention is that more participation and representation of social interests could be a positive asset, but the creation of a regional parliament is not the only way to foster this objective -again, at this stage . One of the main liabilities of Mercosur is, rather than the democratic deficit, the implementation gap : approximately two thirds of the norms sanctioned at the regional level need to be internalized by all the member states to come into force, and by 2002 more than half of them were not (Bouzas, Veiga and Torrent, 2002). A further problem is that the bloc lacks a bureaucracy able to monitor and enforce regional norms. Therefore, it is up to the national bureaucracies to do so, but they lack the incentives to perform such a role. The realists warn not to let form trump function : as appealing as it may be, creating an agency does not guarantee that the job is done. Some necessary tasks are to be performed, but the way to do it may vary across areas and over time . For example, monetary coordination may be desirable, but a common currency is not indispensable -and could even be counterproductive-before the common market is in place . A similar reasoning can be applied to such sensitive issue as democratization: proposing that the citizens control electorally an intergovernmental --as opposed to supranational--organization is contradictory at its best, as they already elect the principals of the organization . For those who advocate deeper integration, transparency rather than democracy should be pursued. The rationale is that the integration process will be enhanced if the initial tasks and initial delegation of authority get to attract the attention of nonstate interests and to provide incentives for them to form transnational interest associations or social movements, which will in subsequent stages demand for better access to regional deliberation and decision-making.
The realists are also wary about the possibility of establishing a common social and developmental policy along the lines of the European structural and cohesion fund model . Such a proposal conceives the bloc as an instrument of social solidarity that not only creates but also redistributes wealth, and is usually accompanied by the endorsement of greater participation of sub-national entities, sectoral representatives, and civil society. The idea, most realists believe, is hardly practicable for two reasons: structurally, because of the heterogeneity of member countries, and instrumentally because there is no regional budget . Heterogeneity does not mean that wealthy and poor countries coexist, but rather that total wealth does not coincide with per capita wealth. Brazil is the largest economy in the bloc and the only possible regional paymaster, but its GNP per capita is lower than Argentina's and its poverty and inequality rates much higher. It would be difficult for Brazilian authorities to explain to their poor that the Mercosur social dimension calls for them to subsidize richer countries. The second challenge is not insurmountable, but as it is rarely addressed it is not likely to be overcome soon . Mercosur lacks a real budget except for the recently created Structural Fund, which will handle at best $100 million a year since 2008 . This amount represents less than 0 .1% of the EU budget and just 0 .01% of Mercosur GDP -whereas the European structural funds are about 0 .4% of the EU GDP (Secretaria del Mercosur, 2005).
It has been pointed out that the progressive proposals usually replicate the EU experience but neglect the process that led to its creationi.e. the Monnet method, based on the perception that regional integration is supported by concrete projects and not by general declarations of basic principles (Malamud, 2005b) . They also fail to recognize that interdependence is a prior condition for successful integration, at least according to the main contemporary theories developed to explain European integration (Moravcsik, 1998 ;Sandholtz and Stone Sweet, 1998). The main driving forces of European integration are transnational transactors, the national governments, the European Commission, and the European Court of Justice (Sandholtz and Stone Swee, 1998), some both an engine and a product of integration . In Mercosur transnational transactors are weak because interdependence is low (currently one eighth of the level of the EU), and there are no supranational actors such as the European Commission or the Court ; therefore, national governments are the only driving force left. In this context, demand for and supply of further integration only emerged from the national chief executives (Malamud, 2003(Malamud, , 2005a. The consequence is that spillover in crucial areas has not taken place, as there is neither delegation nor pooling of decision-making that could enlarge the scope or increase the level of the mutual commitments . In the absence of spillover, the joint political will of the national presidents is what has kept Mercosur going. Mercosur was initially a realistic process that sought gradually and pragmatically to overcome the integration fiction syndrome and the rhetoric that had hitherto predominated in Latin America (Campbell, 1999) . As interdependence increased and economic growth lubricated the process, a pragmatic approach prevailed . After the completion of the schedule of automatic tariff reduction, however, the onset of economic crises ignited intra-regional conflicts and national leaders opted to up the rhetorical ante. Deepening, enlargement and institutional upgrading became obsessive topics as trade controversies multiplied and temporary barriers were erected and taken down time and again . The gap between the progressive and the realists became increasingly wider . Whereas the former talked of creating supranational institutions, a regional parliament, a single currency and a common external policy, the latter underscored the need to promote the liberalization of services and government procurement, to facilitate the process of internalization of regional rules, improve macroeconomic coordination and enhance the decision-making mechanisms without necessarily going supranational . Although there are areas of agreement between the two camps (the need to invest in physical integration and to negotiate in multilateral trade fora as a united group), disagreements far exceed points of consensus. Cognitive dissonance may be a common dysfunction in large processes of regional integration, as the current attempt to constitutionalize the EU shows. In Latin America, however, the phenomenon is reinforced by a historical propensity towards magical realism and high rhetoric. Mercosur was created to mitigate such tendencies, but it has been lately drifting away from "real" realism and back to magical realism.

V. CONCLUSION
Regional integration in Latin America faces unfavorable conditions, although not as harsh as in the past . The triumph of political democracy and market economy throughout the continent has turned integration feasible, although many structural restrictions remain . A major pitfall continues to be the poor level of economic interdependence, be it related to trade or investment (cf. Burges, 2005). Although it has risen in recent years, the figures still are extremely low. A second setback is the closure of most domestic economies: integration means, by definition, tearing borders down and opening up to the neighbors. This has seldom occurred, and public opinion as well as businesspeople in many countries continues to see protectionism as a valid defense against the threat of international competition. Last but not least, the largest Latin American economies have not undergone a process of growing complementarity; on the contrary, they remain frequently competitive both regarding products and markets.
Besides the structural dimensions mentioned above, a series of policy factors lay behind the difficulties at promoting integration. Although in recent years macroeconomic policies have somewhat converged in key countries such as Argentina and Brazil, this was due not to policy coordination but to the unplanned simultaneity of the economic turbulence undergone in those countries. Recurrent political conflicts have not been uncommon either: Venezuela's Bolivarian revolution, Argentina's rambling foreign policy, and Bolivia's new approach to foreign investment in the energy sector have distressed even more the traditionally tumultuous regional landscape.
Given the resiliency of both structural and policy obstacles to integration, regional attempts at regional integration would need to produce an institutional configuration able to enforce minimum collective disciplines while accommodating to national heterogeneity and a large degree of political uncertainty . However, the institutional architecture of most Latin American blocs has failed to deliver: not only was it unable to overcome pre-existing problems, but it has even turned itself into an additional obstacle to further integration as its dysfunctionality contributed to delegitimizing the whole enterprise.
All this said, we do not believe that regional integration in Latin America is doomed to failure. Our argument can be summarized in two points. First, regional attempts have been burdened with unrealistic expectations, against which any outcome would always be judged as failure. Second, they have recurrently been endowed with the wrong institutions, further diminishing their already moderate prospects . The endeavor of integrating countries as badly interconnected as the Latin American ones should start by fostering interdependence through simultaneous economic growth, not by feeding a common identity, creating a supranational community or establishing a joint foreign policy . If the European experience teaches something as concerns regional integration, it is that politics comes much later than economics -if at all (Malamud and Schmitter, 2006) . Interests, not identities, are the fuel that drives these processes. Common institutions, be them intergovernmental or supranational, should acknowledge this fact or risk irrelevance.

Notes
Neofunctionalism is a theoretical approach to the study of regional integration first proposed by Ernst Haas (1964) in the 1950s and 1960s.