Paper for delivery at the 2012 Congress of the Latin American Studies Association (LASA)
Panel “Waves of change in Latin America. History and Politics”
San Francisco, 25 May 2012
Contrary to Europe, where a single process of regional integration has experienced several waves of enlargement, Latin America is characterized by a succession of four waves that saw the signing of agreements launching or reactivating several distinct but quite similar integration processes, in the years 1950-1960, 1970-1980, 1990 and 2000-2010. Most scholarly efforts have been centered on the evolution of each regional integration process in Central America, the Andean or Caribbean regions or in the Common market of the south (MERCOSUR), or on the overall picture of regionalism, yet the simultaneous onsets, the similar features and parallel evolutions of various regional integration processes in Latin America have not been properly studied. This piece fills that void explaining the different waves by a combination of convergence of interests and diffusion of ideas, with a mix of external and internal incentives, in a given historical context. It also puts the emphasis on paradigm shifts intersecting with disruptions, as triggering the surge of a new wave, and it uses path-dependence arguments to consider legacies and resilience. The paper argues that the current wave, as compared to the previous ones, is composed of regional integration processes of a third kind, best described by a contentions blend of structuralism and neoliberalism.
Contrary to Europe, where a single process of regional integration has experienced several waves of enlargement, from its original six member States in 1957 to a total of twenty-seven in 2012, Latin America is characterized by a succession of waves that saw the signing of several agreements launching or reactivating several distinct integration processes. If by wave, we mean a historical sequence during which a similar evolution takes place simultaneously in a given set of countries, then Latin America has gone through four waves of regional integration, weaving a complex patchwork quilt (Table 1).
Table 1. Waves of regional integration in Latin America
Organization of Central American States
Central American Common market
Latin American Free Trade Association
Special Latin American Coordinating Commission
Caribbean Free Trade Association
Eastern Caribbean Common Market
River Plate Basin Treaty
Latin American Economic System
Latin American Integration Association
Organization of Eastern Caribbean States
Common Market of the South
Central American System of Integration
Association of Caribbean States
Group of three (Colombia, Mexico, Venezuela)
Initiative for the integration of Infrastructure in SA
Source: Author’s elaboration.
In the aftermath of WWII, the United States framed the Inter-American system with two important treaties, regarding military (Inter-American Treaty of Reciprocal Assistance, 1947) and political affairs (Organization of American States, 1948). Then the Central Americans created their own sub-regional organization (ODECA, 1951). After this initial, very much cold war-led political step, the 1960s marked the beginning of economic integration in Latin America. This first wave (W1), typified by Rosenthal (1991) as voluntarist, followed the United Nations Economic Commission for Latin America and the Caribbean (ECLAC)’s recommendations to use regional economic integration as a device to promote the industrialization of the continent. This developmentalist and structuralist approach inspired a regional initiative, as well as several sub-regional ones. In 1960, a treaty was signed in Montevideo, given birth to the Latin American Free Trade Association (LAFTA)1, and the same year, the Central Americans launched their Common Market. Later, the Caribbean and Andean countries would do the same. For all, the European Economic Community (EEC) was a source of inspiration. The decade closed with a treaty inviting Argentina, Brazil, Uruguay and Paraguay to “combine their efforts for the harmonious development and physical integration of the River Plate basin”, reflecting strategic concerns of the military regimes.
During the 1970s and 1980s, a widespread disappointment over the scope of trade liberalization and the degree of industrialization, led the promoters of integration to downgrade their goals and give up tight schedules. The Latin American Integration Association (LAIA) replaced a stalemated LAFTA. In the Caribbean region, new organizations also replaced older ones. Following the example of the River Plate treaty, an agreement was reached between the Amazonian countries. Finally, two organizations (SELA and Rio Group) were created to provide forums of consultation for economic or political matters. This second wave was, hence, a “revisionist” one (Rosenthal, 1991).
During the 1990s, regionalism was framed by a new dominant ideology: neoliberalism. The so-called Washington consensus was mute about regional integration, yet neo-classical economics envisions free trade and free markets as the only ways to foster growth. Even the ECLAC revisited its doctrine, pressing for “open regionalism” (ECLAC, 1994). Stimulated by U.S. trade initiatives (Initiative for the Americas, North American Free Trade Agreement, Free Trade Area of the Americas), and by the European steps toward a unified market, a new treaty was signed in South America (MERCOSUR, 1991), while older ones were renegotiated in Central America (SICA, 1991), the Caribbean (ACS, 1994), and the Andean region (CAN, 1996).
Finally, as the neo-liberal era came to an end, and as the continent massively turned to the left (Weyland, Madrid & Hunter, 2010; Cameron & Hershberg, 2010; Levistky & Roberts, 2011; Dabène, 2012b), trade driven integration was object of strong criticism. While the United States signed bilateral agreements, as substitutes to the ill born Free Trade Area of the Americas (FTAA), and some initiatives were still motivated by a will to facilitate trade (IIRSA and Pacific Arch), a new conception of integration emerged. Venezuela-sponsored ALBA includes Trade for the Peoples Treaties accompanied by supplies of cheap oil. UNASUR and CELAC have a rich agenda of functional cooperation, ranging from defense and security to infrastructure and environment.
How can we explain this succession of four waves? How can such a complex and irregular pattern of evolution be explained? Why so many similar agreements signed successively or simultaneously in different sub-regions? How different is the last wave from the previous ones? These are the questions raised in this paper.
Theoretical framework, hypotheses and approaches
To answer these questions, the literature on regional integration in Latin America is not much of help. Using different theoretical perspectives, there are significant scholarly efforts to explain why and how some treaties are signed in a given historical context, from Schmitter (1970a) to Mattli (1999) or Duina (2006), among others. There are also excellent accounts of how they evolved through time, in the whole continent (Bawa, 1980), or in some particular sub-regions (Cohen, 1972; Wynia, 1972; Lewis, 2002). The discontinuous pattern of evolution has also been studied. Schmitter (1970b), for instance, described crises leading key actors to reevaluate the level and the scope of their commitments to regional institutions. In the same vein, Dorette Corbey (1995), referring to European integration, also described a “stop and go” pattern of evolution, that owed to the governments meeting pressure groups’ demands for protection (stop phase), but then realizing the costs of rivalry are high (go phase). More recently, Dabène (2012a) offers an explanation referring to cycles of politicization, depoliticization and repoliticization.
These works help understand why regional integration processes are likely to resemble bumpy roads, yet they do not analyze the simultaneity of launchings or relaunchings of similar regional integration processes. With the exception of Rosenthal (1991), insufficient attention has been paid to the waves of agreements, probably because the literature always has in mind the European case, with its unique progressively enlarged process. The simultaneity and similarity of various regional integration processes remain to be explained and this is one of this paper’s empirical and theoretical objectives. The other objective of this paper is to offer a reflection on the last wave.
How can we explain a wave of integration?
I argued in a previous work that waves of political evolution could be explained by a combination of convergence and diffusion (Dabène, 1997). Waves of regional integration are no exceptions. Convergence stems from the initial economic and political environment. In distinct groups of countries, the same interests are prompting integration and the governments collectively address the same development issues in similar ways. Different groups of countries, pressed by interest groups, may be simultaneously convinced that regional economic integration is the proper tool to foster growth and promote development because they all are influenced by the same policy paradigm (Hall, 1993). This convergence can be the byproduct of a diffusion of ideas or norms. In that case, diffusion entails convergence. But it can also be more spontaneous. Economic crises act often as critical junctures (Collier & Collier, 1991) generating similar reactions. Keen to develop the same projects, under the influence of the same ideology and proactive actors, different groups of countries set similar agendas and try to deliver the same outcomes. Failing to do so, they react the same way, but with distinct intensities, depending on the lessons learned by the actors, the new global economic environment and the political coalition holding power in the dominant countries.
In parallel, some models of integration are emulated. The model that triggers the spread of integration can come from within or outside the region. Latin America offers a study case of confrontation between domestic and imported models. The continent has elaborated its own theorization of integration, but at the same time the European Union and the United States have tried to influence it. Much has been written about the European influence (Smith, 1995; Mattli, 1999; Medeiros 2000; Grugel, 2004 and 2007; Dri, 2011). Indeed, the European construction has become an inevitable point of reference, and a mechanism of cognitive shortcut (Weyland, 2006) has been active in the region. Yet the European Union never managed to fully impose its conception of a deeper integration.
The objective of this paper, as previously mentioned, is to explain parallel births and rebirths of integration processes in Latin America. The distinction between “birth” and “rebirth” is important, and this paper will borrow from the historical institutionalism tool kit to address it. We know how sticky and resilient institutions can get (Pierson, 2004), and a “rebirth” can never be a complete across the board change. Some points made by historical institutionalists studying European integration cannot apply to Latin America, simply because no regional integration processes has crossed the threshold of supranationality and member states, or more precisely Presidents (Malamud, 2003) remain key actors. Regional integration processes in Latin America have generated path dependence, but the institutional arrangements have not yielded increasing returns. Yet other notions are very useful. Some gaps (Pierson, 1996) can emerge between governments’ policy preferences and the outcomes delivered by the regional institutions. As a result, even though regional institutions evolve through time, for instance through layering (Thelen, 2003), some countries can simply decide to sign new agreements.
The purpose here is not to fully examine the parallel evolution of all Latin American regional integration processes, leading to growing gaps and subsequent relaunchings, a task I tentatively and partially fulfilled elsewhere2. Leaving aside some core features of this evolution, such as the diffusion of institutional designs or redesigns or policy transfers, some evolutionary factors will be taken into account when they help explain the surge of a new wave. Among them, paradigm shifts are central. They have a transformative capacity, by reframing the actors’ representations, projects and agendas. Yet they cannot be a sufficient condition for launching a new wave. They have to intersect with some sort of disruption.
We therefore have two main hypotheses to explain the waves of integration:
Regarding the first wave (W1):
H1: W1 is best explained by a combination of convergence of interests and diffusion of ideas, with a mix of external and internal incentives, in a given historical context. As for the next waves:
H2: Wn is best explained by a paradigm shift intersecting with a disruption, in addition to a new combination of convergence of interests and diffusion of ideas, with a mix of external and internal incentives, in a given historical context. When Wn is a transformation of Wn-1, given classical path-dependence arguments, we can add a specific hypothesis:
H3: Wn is best described as a contentious wave, with new elements competing with resilient ones. The other objective of the paper is to closely look at the current wave of integration. The previous waves of integration have been successively framed by two paradigms, structuralism and neoliberalism. The new one could either swing back to structuralism, either be inspired by a new paradigm. Building on the discussion of the first three waves, I will argue that during the years 2000-2010, Latin America has ventured into new territories. A regional integration of a third kind is under way, combining and synthesizing several features of the previous waves. Wave 4 is not exclusively centered on trade like the third one, nor is it protectionist like the first one. As a result, W4 is not as homogeneous and compact as the previous waves. Our last hypothesis supplements H3, looking at the content:
H4: W4 is an integration of a third kind, a contentious blend of structuralism and neoliberalism. In order to test these hypotheses, the paper mainly uses an ideational approach. Each project of integration is shaped by a dominant paradigm, and a new wave is often signaled by a paradigm shift. The emphasis is put on the key actors’ representations, intentions, projects and agreed-upon agendas, and the way they can be shaped by norm diffusion. As mentioned earlier, the European Union is an important actor seeking to export its “model” in Latin America, competing with the United States. The EU has not being equally convincing in all sub-regions, due to a series of factors such as political and economic leverage, or the misfit between European norms and local ones. Indeed, there are different degrees of localization (Acharya, 2004).
As important as ideas can be, they cannot exhaust the explanation of waves. In order to fully understand why there are several distinct processes launched or relaunched at the same time, the analysis will also stress interest driven behaviors.
The paper also uses a process-tracing approach, in order to unveil the foundations of the decisions that led to the signing of agreements. Trying to explain the outcomes, i.e. the main features of the waves and in particular the last one, the paper goes back in time and offers analytic narratives that connect the actors’ intentions to agendas and outcomes. In the following sections, I proceed to compare sequences of events constituting the waves and show that they share commonalities. I assess the exogenous influences, such as the European or the North American ones, looking at the way their own historical sequences eventually overlap with the Latin American ones and showing that the timing of events matters a great deal (Mahoney & Rueschemeyer, 2003). And I also evaluate the domestic politics factors, using an “inside out” framework (Jayasuriya, 2003). The last wave, in particular, is clearly driven by a political swing in the region.
The case narratives are rather synthetic for the first waves. As regards the current wave, the presentation is more systematically organized around the following lines: paradigm, actors, agendas and methods, elements of convergence and diffusion, and external influences3.
Explaining the first waves W1
Following the release of a report in 1949 by Argentine economist and ECLAC’s emblematic general secretary, Raúl Prebisch, Latin America was offered a roadmap for its development strategy. Titled “The economic development of Latin America and its principal problems”, this seminal work, Hirschman (1961) referred to as the ECLAC’s manifesto, laid the basis of the unequal exchange theory and sparkled a paradigm shift in a region where the theory of comparative advantage had long been popular, at least until the Great Depression. Prebisch theory was grounded in empirical observations and professional practice as Argentina’s Central Bank general manager from 1935 to 1943 (Dosman, 2008). In the aftermath of the Depression, Argentine exports revenues soared and industrialization became a hard necessity (Love, 1980). It became clear that export-led growth was no longer viable and he started advocating for inward-looking development and industrialization as a way to reduce the vulnerability of the Latin-American economies. Furthermore, Prebisch asserted that unification of markets could yield productivity gains and accelerate industrialization, provided it was protected by high tariffs. The recommendation was to launch a strategy of import-substituting industrialization (ISI) on a collective basis.
It is important to bear the timing in mind. Prebisch’s proposals were released at the beginning of the 1950s, a period marked by the Korean War, that saw a rise in prices paid for Latin American commodities in the world markets. In this context, the pessimistic theory of unequal exchange could hardly convince Latin American policy makers. However, it would not last very long before the terms of trade declined again for Latin America. In addition, the U.S. opposed the creation of ECLAC right from the beginning, arguing that it duplicated the function of the Inter-American Economic and Social Council (IA-ECOSOC), and was also very fierce critics of the ISI strategy.
This rather adverse initial environment did not prevent ECLAC from opening a sub-regional office in Mexico City in 1951, and started lobbying in Central America. Why this region? Central America had a long history of failed attempts to resuscitate the Federation that lasted twenty years after the independence. After WWII, Guatemala and El Salvador were active trying to convince their three neighbors to sign political pacts. They eventually managed in 1951 to create the Organization of Central America States (OCAS), soon to be paralyzed by rivalries between revolutionary Guatemala and the rest of the region. Interestingly, whereas the political project failed, ECLAC started in 1952 to organize meetings in the framework of a Central American Economic Cooperation Committee. Three Central American economists initially presented the idea of economic integration during ECLAC’s fourth session in México City in June 1951. They were keen to insulate their project from the tense political climate prevailing in the isthmus (Urquidi, 1998). ECLAC quickly endorsed it and tried to sell it to the governments of the region. They met with different reactions, ranging from enthusiasm in Guatemala and El Salvador, to indifference in Nicaragua and hostility in Honduras and Costa Rica, where the banana exporters, mainly U.S. companies, were strong supporters of free trade (Sáez, 2009). Nonetheless, the project of a Central American Common Market (CACM) was on the agenda, and ECLAC was very active promoting it.
In 1958, a first Multilateral Treaty of Free Trade and Integration was signed, followed by a more encompassing one (Treaty of Managua) in December 1960 creating the CACM. The former limited free trade to a short list of products, while the latter was more compatible with the U.S. position defending free trade for all products with eventual exceptions. The U.S. also managed to have the Central Americans give up their project of having integrated industries, contending that it would create monopolies. Until the last moment, they also tried to convince, most notably El Salvador and Honduras, to sign a free trade agreement, instead of the ECLAC sponsored treaty (Urquidi, 1998). The U.S. at that time had an inter-American agenda focused on security matters, meaning that their purpose in the economic field was to “gain acceptance of Latin America that the defense program is vital to both” (OAS, 1951). Their main concern was to secure imports of raw materials.
The Central American sequence of integration ended abruptly with a war between Honduras and El Salvador in 1969. Retrospectively, an important actor admitted that they had acted as bureaucrats (técnicos), without considering the importance of the political factors (Urquidi, 1998). This depoliticization of integration (Dabène, 2012a) proved fatal.
In the rest of the continent, four series of reasons converged to propel regional integration.
First, by the end of the 1950s, even the countries where export-led growth had prevailed were ready to adopt the ISI model (Bulmer-Thomas, 2003). Second, a new generation of statesmen emerged, for whom regional integration was a noble political quest: Arturo Frondizi (Argentina), Juscelino Kubitschek (Brazil), Romulo Betancourt (Venezuela), Alberto Lleras Camargo (Colombia). The latter had previously served as first General Secretary of the OAS, during the crucial years (1948-1954) when the U.S. was harassing ECLAC, and had a good working relation with Raúl Prebisch. Third, the U.S. reluctance softened. While the U.S. administration refused a Chilean project to create a development bank in 1954, or paid little attention to a Brazilian proposal of a Pan-American Operation in 1958, they finally agreed in April 1959 to create the Inter-American Development Bank (IADB), a major step to provide much needed financial support to integrated projects in the 1960s. The Cuban revolution triggered this policy change, epitomized by Kennedy’s Alliance for progress (1961), which strongly supported regional integration. And fourth, the first steps of the European construction and the success of Central America came as additional supports to the regional integration advocacy network in the continent.
In February 1960, the Montevideo treaty was signed, creating the Latin American Free Trade Association (LAFTA). The reasons why LAFTA never achieved its objective to eliminate intra-regional tariffs have been analyzed (Bulmer-Thomas, 1997). Suffice is to say, for the sake of explaining the waves of integration, that the issue of gain distribution was central for less-developed countries (Bolivia, Ecuador, Paraguay and Uruguay).
The idea of a separate agreement between less-developed countries was first suggested in 1964 during a LAFTA meeting, and then it was object of a Declaration signed in Bogota in 1966 by the presidents of Colombia, Chile and Venezuela. Some projects of mutual interests were discussed and an Andean development corporation (CAF) was created in 1968 to finance them. The idea of temporary sub-regional agreements between less-developed countries received a decisive backing during the 1967 Punta del Este summit of American presidents. In the final declaration, the Presidents agreed to “promote the conclusion of temporary sub-regional agreements, with provision for reducing tariffs within the sub-regions and harmonizing treatments toward third nations more rapidly than in the general agreements, in keeping with the objectives of regional integration”. But when the Andeans4 signed the Cartagena Treaty in 1969, they went far beyond putting in place a temporary device for reducing tariffs. The Andean Pact was very ambitions, targeting a customs union and aiming at harmonizing economic and social policies. The integrated industrial program and the foreign investment codes were centerpieces of their ISI-inspired project (Parkinson, 1973). The Andean Group (GRAN) was also granted a supranational institutional arrangement very much inspired by the European one, with a Commission making majority decisions, and a Secretary (Junta) composed of three members representing the regional interests and not taking orders from their respective governments.
Table 2 summarizes the main features of the first wave. Two points deserve to be highlighted.
First, W1 has been activated by a paradigm shift that, at some point, and thanks to ECLAC’s técnicos (Wynia, 1972), convinced policy makers that the center/periphery theory matched with economic realities. Then three historical sequences overlapped, in South America, Central America and Europe, producing a mix of convergence and diffusion. Another key actor, the United States, offered resistance. As a result, W1 is a contentious process. The paradigm shift was incomplete and not hegemonic, as the U.S. (and the International Monetary Fund) kept on pressing for export-led growth. The first two agreements, CACM and LAFTA, felt short of ECLAC’s expectations in terms of industrial planning. Being a latecomer, at a time when U.S. pressures weakened and the shortcomings of LAFTA and CACM were visible, GRAN was much more in tune with ECLAC’s doctrine.
Two, the proliferation of agreements and subsequent waves were not meant to be. The 1967 Declaration of the Presidents of America reads: “Presidents of the Latin American Republics resolve to create progressively, beginning in 1970, the Latin American Common Market, which shall be substantially in operation in a period of no more than fifteen years. The Latin American Common Market will be based on the complete development and progressive convergence of the Latin American Free Trade Association and of the Central American Common Market, taking into account the interests of the Latin American countries not yet affiliated with these systems”. Moreover, the U.S. President declared “his firm support for this promising Latin American initiative”. In the 1970s, the merging process never got started. The sub-regional processes entered a “zone of indifference” (Schmitter, 1970b) and survived by default, but not without generating path dependency.
W1 lost steam at the end of the 1960s and got completely deadlocked in the 1970s for a number of reasons that cannot be exposed here in details. Some have to do with the way regional integration was conceived in the first place, with inefficient mechanisms of negotiation, weak attention paid to gain distribution issues and institutional flaws. Others remit to domestic politics, with some countries experiencing profound changes, such as Chile in 1973. Voices were also loudly heard criticizing foreign companies domination of the free trade areas (Wionczek, 1970). And finally, the international context changed, with the end of the Bretton Woods system in 1971 and the oil shocks of 1973 and 1979. According to Bulmer Thomas (2003: 299), “the procyclical nature of intraregional trade was a disappointment for those who had hoped that regional integration would increase the autonomy of the region in the face of external shocks”. On a more political note, Prebisch (1969) lamented that nationalisms had not been amalgamated into a ”Latin-American nationalism” that could have cemented development on a collective basis.
In the Andean Group, Chile was the troublemaker. The Chicago boys surrounding General Pinochet were pressing for changes. They were keen to reduce tariffs and attract foreign investors. They managed to have the GRAN introduce modifications, extending deadlines for the Common external tariff approval among others. These provisions introduced flexibility, without meeting Chilean demands. It withdrew from the group in 1976. Other Andean countries, such as Bolivia, Colombia and Peru, experienced drastic political changes that led their government to favor the trade liberalization model, instead of the ISI one. They eventually downgraded their commitment to GRAN (Avery, 1983). In Central America, the war between Honduras and El Salvador had devastating effects, Honduras putting an embargo on transit trade between El Salvador and Nicaragua and Costa Rica.
Displeased with regional integration and ISI, many countries emulated Chile and started to turn to export promotion. Even ECLAC learned its lessons. The failure of the 1950s ISI model (rising inflation, increasing balance of payment deficits) and the paralysis of regional integration processes had its experts try to amend the doctrine. It was all the more necessary because, once again, ECLACS’s theory was contradicted by realities. During the 1970s, the terms of trade got very favorable for Latin America’s commodity exporters. ECLAC did not, however, have a clear alternative, so they proceeded to adjust the paradigm. This second order change (Hall, 1993) entailed more flexibility and new instruments. The new Latin American Integration Association (LAIA) was a loose organization with no ambition other than to legally register partial scope agreements signed by its member states.
ECLAC’s loss of credibility allowed another initiative to see the light. In 1975, all twenty-five Latin American and Caribbean countries created the Latin American Economic System (SELA) in order to promote regional cooperation and establish a permanent system of consultation. SELA replaced the moribund Special Latin American Coordinating Commission (CECLA) created in 1964 initially to set common Latin American positions in the UN Conference on Trade and Development (UNCTAD). Contrary to LAFTA, both SELA and LAIA gave “prominence to industrial complementarity and private sector initiatives” (Bulmer Thomas, 2003).
SELA was not born out of a paradigm shift for regional integration, but it did reflect a growing concern about Latin America’s need to speak with one voice to the United States. Originally a Mexican-Venezuelan proposal, it was also influenced by their affinity with the non-aligned movement. In a context of fierce bi-polarization, and with repressive military regimes in the southern cone keen to have close relations with the United-States, Latin America was politically fragmented. During the 1970s, Milenky (1977) was right to point out that disintegration had occurred, while at the same time interdependence had increased.
The beginning of the 1980s did not bring about significant changes. When the debt crisis erupted in 1982, Latin America proved to be very vulnerable. Economic integration was no longer on the agenda. However, the 1980s saw the progressive democratization of the continent, and some sense of solidarity resurfaced. In 1983, Mexico, Venezuela, Colombia and Panama, soon joined by Argentina, Brazil, Peru and Uruguay (Contadora Group), offered mediation to a war-ravaged Central America. And in 1984, Mexico, Argentina, Brazil and Colombia organized a summit in Cartagena (Colombia) to discuss the debt issue. Both diplomatic initiatives epitomized a new spirit of collaboration between recently installed democracies. The Contadora Group’s proposal failed, but the governments involved in the effort decided in 1986 to create a permanent mechanism of consultation. This repoliticization (Dabène, 2012a) laid the ground for the third wave of integration.
Turning to the external actors during the period, the panorama is complex. The U.S. was very discrete, at least regarding regional integration. Following the 1967 summit, the Latin Americans used CECLA to collectively bargain with the U.S. That effort proved disappointing and CECLA was abandoned after 1973 (Milenky, 1977).
Conversely, the EU was very active trying to export its model of integration. In the Andean region, the EU secured technical assistance for institutional building, targeting the JUNAC. Nonetheless, some Latin American exports got hurt by the European common agricultural policy, and there was no inter-regional institutionalized dialogue to find common ground, other than the meeting of Latin American ambassadors in Brussels who created a coordination group. The 1975 Lomé convention further affected South American interests. During the 1970s, the EU signed several bilateral agreements, showing that a common policy for Latin America had not yet been thought of, other than offering technical institutional cooperation. The years 1983-1984 marked a change, with the EU establishing formal agreement with the Andean and the Central American regions. When Portugal and Spain joined the EU in 1986, this embryonic interest for Latin America received a decisive boost.
Table 3 summarizes the main features of W2.
By all account, W2 looks more like a trough of a wave than a real wave. It was a long intermediate and quite inactive period between two paradigm shifts. The private sector played an important role, preserving some level of intra-regional trade in times of economic nationalism, but all the integration processes became deadlocked. Even the Caribbean Community (CARICOM), created in 1973, was considered a major disappointment (Payne, 1981). Still, because it lasted almost two decades, and because the EU actively helped strengthening the institutional arrangements, W2 turned out to be even more path dependent. The regional institutions were object of political indifference, but that left them room for being autonomous and assertive. As a consequence, during W2, they enhanced their resilience and stickiness (Pierson, 2004) but also in some cases their institutional layering. Political indifference was the product of a growing gap between policy preferences and integration outcomes. Important disruptions during the period (economic crises and transitions to democracy) did not intersect with a paradigm shift, preventing a massive relaunching of a new wave.
Table 3. Main features of Wave 2
Second order change to developmentalism, end of consensus
Private sector, EU
Flexible planning with no fixed deadlines
Lessons learned from ISI shortcomings
Chilean (and Asian) model of export-led growth
Group of Latin American Ambassadors
Bilateral agreement with Argentina
Bilateral agreement with Uruguay
Bilateral agreement with Brazil
Bilateral agreement with Mexico
Lomé Convention with Asian Caribbean & Pacific countries
Cooperation agreement with Andean region
San José dialogue with Central America
Single European Act
For the reasons above mentioned, the second half of the 1980s witnessed a renewed integration spirit in the region. The recently democratized countries were keen to work together to deliver some regional goods. Brazil and Argentina started in 1986 with a bilateral agreement seeking to initiate a cooperation driven by a will to bolster mutual trade and collectively defend democracy (Dabène, 2009; Gardini, 2011). The 1986 Argentina-Brazil Economic Integration Pact (ABEIP) was still influenced by ECLAC’s structuralism, as it targeted intrasectorial complementarity, technological modernization, and harmonization of economic policies. Yet, it was more realistic and pragmatic than previous integration schemes (Manzetti, 1990). ABEIP was eventually supplemented by two dozens protocols, signed on a bilateral and then on a trilateral basis with Uruguay.
The paradigm shift occurred at the beginning of the 1990s simultaneously in all sub-regions, and coincided with changes in the global economy and with policy swings in the U.S. and Europe (Manfield, Milner, 1999). Again, a mix of convergence and diffusion was perceptible. Under U.S. hegemony, neo-liberalism was the new dominant ideology at the end of the Cold War, and with it a new trade-centered market-friendly conception of integration influenced the signing of new agreements and the evolution of older ones, in Latin America as well as in the rest of the world5.
In South America, the elections of three new presidents, Carlos Menem in Argentina, Luis Alberto Lacalle in Uruguay and Fernando Collor in Brazil, between December 1989 and March 1990, meant a clear turn to the right. When Argentina, Brazil, Paraguay and Uruguay signed the Asuncion treaty giving birth to MERCOSUR in March 1991, their purpose was not to promote economic development, but rather to liberalize trade and adopt a common external tariff. In the Andean region, after the 1987 Quito Protocol introducing flexibility regarding policy harmonization, the member States adopted in 1989 a new trade-centered agenda and went on converting their group into an Andean Community (CAN) in 1996. In Central America, after a sequence of reactivation following the pacification of the region and the creation of SICA, the integration process was put on a new track in 1997 when the region decided to implement the reforms recommended by the IADB and ECLAC (1997).
It is worth noting that for the first generation integration processes, the paradigm shift was incomplete and contentious. The Andean and Central American regions both adopted a new neoliberal agenda in the 1990s, but it hardly tricked down to all regional agencies. CAN and SICA were complex multilayered institutional arrangements with many bodies keeping traces of the paradigm that was dominant when they were created.
During this period, ECLAC also experienced the paradigm shift, embracing “open regionalism”. According to ECLAC (1994), “What differentiates open regionalism from trade liberalization and non-discriminatory export promotion is that it includes a preferential element, which is reflected in integration agreements and reinforced by the geographical closeness and cultural affinity of the countries of the region”. As in Asia (Jayasuriya, 2003), open regionalism entailed a political project centered on regional governance. Contrary to what institutional arrangements of the previous waves provided, open regionalism was supposed to be run by flexible and informal rules. It clearly reflected the new political and economic power of the tradable globalized sectors and the neoliberal coalitions winning elections during the 1990s.
On the external front, Latin America was submitted to pressures from the U.S. and the EU, representing partially competing modes of governance (Grugel, 2004).
U.S. President Georges Bush launched his Enterprise for the Americas Initiative (EAI) in June 1990, emphasizing free trade, investments and debt relief. The policy initiative was designed to help “virtuous” countries, locking in the neoliberal reforms implemented. To qualify for debt reduction funds, a country had to have signed a standby agreement with the International Monetary Fund and had to receive a structural adjustment loan from the World Bank (Graham, 1991).
NAFTA negotiations started in June 1991 as the first free trade agreement envisioned by EAI. In parallel, all Latin American countries, except Cuba, Haiti and Surinam, signed framework agreements as preludes to free trade negotiations with the U.S. NAFTA clearly set new standards for trade negotiations in the region. Building on the 1988 Canada – U.S. Trade Agreement (CUSTA) (Whalley, 1992), NAFTA introduced new issue areas and a new discipline, going far beyond what was on the WTO’s agenda. NAFTA extended the notion of trade promotion to services, sanitary and phytosanitary measures or to intellectual property rights, and established regulations for government procurements and investments. Furthermore, the negotiations also included some “brand new” issues such as environment and labor rights.
In the aftermath of NAFTA’s coming into force, a mix of diffusion and convergence was set in motion. The U.S. hosted in December 1994 a Summit of the Americas, inviting all thirty-four democratic countries of the Hemisphere, only leaving aside Cuba. Product of a “cascading modular multilateralism” (Feinberg, 1997), the Summit’s agenda was rich and rather consensual. Twenty-three mandates were discussed, grouped around four objectives: preserving and strengthening the community of democracy in the Americas; promoting prosperity through economic integration and free trade; eradicating poverty and discrimination in our Hemisphere; and guaranteeing sustainable development and conserving our natural environment. Yet it was the proposal of a Free trade area of the Americas (FTAA) that captured the attention of both the business community and the governments in the Hemisphere. In parallel, many governments embarked upon a legislation upgrade, anticipating that FTAA discipline would one day be multilateralized, thus using FTAA as a rehearsal in the perspective of future WTO rounds of talks.
Officially opened during the second Summit of the Americas in Santiago (Chile) in 1998, the FTAA negotiations were paralyzed during the Miami ministerial meeting of November 2003, for a series of reasons that need not be analyzed here. What deserves to be underlined is the NAFTA – FTAA effect in Latin America. Some countries emulated this type of agreement. Colombia, Venezuela and Mexico, for instance, signed a NAFTA copy-paste free trade agreement in 2004. Other existing regional blocks were prompted to include in their agendas rules regarding investments or intellectual property rights. In short, NAFTA – FTAA’s agenda and discipline became the norm in the Western Hemisphere. However, it also became the symbol of U.S. pro-market interference in Latin American politics. Mexico, in particular, was object of a NAFTA-ization working through institutionalization, harmonization, communication and persuasion/mimicking (Aspinwall, 2009). Even for Canada, according to Clarkson (2002), NAFTA constitutionalized a pro-market neoconservative ideological orientation that would constrain future policy-making. As a result, the whole FTAA negotiation triggered an intense transnational mobilization that contributed to the transition from Wave 3 to Wave 4 of regionalism. It is worth noting that despite FTAA failure, the Summit process6 kept on working on the non-trade agenda, building a multilevel regional governance scheme (Dabène, 2009).
The European Union was also very active during the 1990s. The EU embraced open regionalism and promoted inter-regionalism, but it also signed agreements with two countries: Mexico in 1997 and Chile in 2002. In 1999, a first European – Latin America Summit was held in Rio, Brazil. It set the tone for the subsequent meetings and negotiations7. The EU pushed for an agenda encompassing trade liberalization, economic and social cooperation and political partnership. Besides the trade agenda, that looked similar to the one the U.S. was negotiating, the EU clearly intended to promote its best practices and its conception of social inclusion (Grugel, 2004).
The EU LAC summits displayed harmony and value convergence, but did not deliver much concrete progress, most notably in the realm of trade facilitation. The inter-regional negotiations ran into many obstacles. As we shall see, during the fourth wave, some negotiations were concluded (Caribbean region, Central America); others were deadlocked (MERCOSUR).
Table 4 summarizes the main features of W3.
W3 was clearly the product of a paradigm shift intersecting with important changes. The 1990s saw the Latin American regional groups embracing neo-liberalism, both because the Washington consensus was actively diffused and because the continent was ready to turn its back to a protectionist model of development that was associated with the debt crisis. This decade led some observers to imagine that the conditions were ripe for the Western Hemisphere to build a community, because of an unprecedented level of convergence of values and interests (Inter-American Dialogue, 1992). The next decade proved them wrong.
Table 4. Main features of Wave 3
Neo-liberalism, open regionalism
Economic crisis, lessons learned
Initiative for the Americas Enterprise (IAE)
Rio Group Dialogue
Inter-institutional agreement (LAC)
Framework cooperation agreement with Central America
Framework cooperation agreement with Andrean region
North American Free Trade Agreement (NAFTA)
Summit process (LAC)
Inter-regional framework cooperation agreement with MERCOSUR
Economic partnership, political coordination and cooperation agreement with Mexico