Blended Diplomacy: Institutional Design and Brazil’s National Interest in Trade

Introduction

Trade negotiations are two-level games (Putnam 1988), implying that diplomats face domestic and international constraints. Policymakers, therefore, must make compromises in the process of elaborating the national interest. In doing so, bureaucrats and elected officials may disagree with each other while facing pressure from organized interests and foreign diplomats. In trying to understand a scenario, policymakers usually resort to ideas (Keohane and Goldstein 1993). However, both elected and unelected officials in charge of crafting a country’s stances in negotiations are embedded in institutional settings (Allison 1971) located in the interstice of the fluid divide between the domestic and the international levels (Chorev 2007). Such a fact prompts a question: how does the diplomatic institutional design impact a country’s views on a trade negotiation? That is, do differences in the manner bureaucracies at the forefront of trade diplomacy relate to the international level and interact with domestic constituencies—including other state institutions—explain the national interest?

I answer those questions through the development of two categories of diplomatic bureaucracies. Such institutions correspond to the section of the state that coordinates policymaking in a negotiation, dealing with domestic inputs (including those from other state institutions) and representing the sovereign state in diplomatic summits. If the diplomatic bureaucracy is responsible for foreign policy in general, it blends politics with market-related concerns, being more embedded in the international level than in the domestic one. Alternatively, if such a bureaucracy is focused on economic issues only, domestic interests become the driving force in shaping negotiators’ mindset. The distinction generates the two ideal types: blended diplomacy, which is more susceptible to foreign factors—including concerns related to high-politics—, and economic-focused diplomacy, which is more sensitive to the domestic dynamic of power—in particular pressures from interest groups. Of course, I do not assume that foreign policy can be conducted without any economic concerns. The point is that blended diplomacy subjects the immediate demands of market actors to a political filter that, in the eyes of policymakers, take into consideration long-term, state-centered goals. While policymakers can deliberately change institutions, the latter evolve in directions not foresaw by their creators (Pierson 2000). Therefore, the design of the diplomatic bureaucracy in charge of trade negotiations has independent consequences upon the formation of the national interest as well as on state-society interactions. Here I build upon Krasner’s (1978, p. 35) notion of national interest, corresponding to general goals that decision-makers set, but not to specific preferences of interest groups or office-holders.

I develop the argument following the model of Beach and Pedersen (2013, p. 17), combining process tracing with the purpose of theory-building. It implies in building first a conceptual framework and later demonstrating its operationalization and empirical application. With such a purpose, I study Brazil’s case during the negotiations of World Trade Organization’s (WTO) Doha Round and its Development Agenda (DDA) between 2003 and 2008. The Brazilian case approximate the blended ideal type as the Ministry of Foreign Relations (MRE, known as Itamaraty) oversees diplomacy in general and defended the liberalization of agricultural markets as a means of enhancing Brazil’s position in international politics and soft-balancing (Hurrell 2006) the dominance of the European Union (EU) and the United States (US) in global economic governance. The counterfactual argument implies that, should Brazil have concentrated the formulation of negotiating positions in the hands of a focused bureaucracy, the national interest would have leaned towards protectionism. I reach the conclusions based upon semi-structured interviews with bureaucrats, office holders, members of business associations, and NGOs, triangulated with archival sources, and secondary literature (Beach and Pedersen 2013, p. 123-143).

The findings suggest that Brazil could have devised a more balanced approach towards the DDA in the case it had not submitted its national interest in trade to broader considerations related to international politics. In contrast to world-wide trends at that time, Brazilian diplomats refrained from seeking second-best options to multilateral liberalization, such as the pursuit of preferential trade agreements (PTAs). Although the DDA is still under negotiation, I delimited the time-frame of the study to control for alternative explanations, thus allowing me to increase the analytical leverage despite working only with a single case. The year 2003 marks the rise of Brazil as one of the main negotiating parts of the DDA. Along with India, the country displaced Canada and Japan from the so-called Quad, also formed by the EU and the US (Patriota, this volume). Furthermore, the failure of concluding the 2008 July Mini-Ministerial Meeting in Geneva represents the main deadlock that the negotiating parts of the DDA faced before the beginning of the global financial crisis (Narlikar and Van Houten 2010, p. 142). Since then, the prospects for effective multilateral liberalization of trade diminished significantly despite the establishment of the Trade Facilitation Agreement after the Bali Ministerial Meeting in 2013 (Rodrigues Vieira 2016).

Apart from being one of the most active negotiating parts representing interests from outside the West in the DDA (Hopewell 2015), Brazil has unique characteristics that make it a crucial case for developing the ideal types outlined above. Brazil’s defense of liberalization is puzzling under alternative frameworks based on ideas, organization of interests, government preferences, and current institutional accounts. Those explanations predict the opposite set of national interests. In Brazil, MRE supposed autonomy vis-à-vis interest groups and historical alignment with the old import substitution industrialization (ISI) paradigm (Sikkink 1991) within a context of strong organization of manufacturing interests and economic debate pervaded by neo-developmental ideas (Ban 2013) would predict the defense of protectionism. Yet, the country remained committed to economic liberalization. Moreover, in the multilateral system of trade, Brazilian diplomacy has historically contested European and US dominance even under the period of the General Agreement on Trade and Tariffs (GATT, 1947-1994), before the creation of the WTO (Narlikar and Tussie 2004).

With the blended/focused typology, I contribute not only to the emerging scholarship that discusses the strategies and foreign policy of emerging powers (Carranza 2016; Milani and Pinheiro 2016), but also to the literature that discusses the weight of state institutions on decision-making related to domestic-international phenomena (Farrell and Newman 2014). This is crucial to understand how states cooperate in a multipolar world (Patriota, this volume). In negotiations, institutions are often considered as being mere intervening factors to channel domestic interests, changing the weight of those interests on the final set of positions a country defends (Ikenberry 1988). Moreover, it remains unclear how different bureaucracies interact with each other, as well as whether different institutional settings lead the “fluid divide” between domestic and international constraints to generate distinct opportunities for a given country, as Chorev (2007) argued based upon the U.S. case.

In the first section, I explore the existing explanations for Brazil’s positions in the DDA negotiations. I consider institutional factors, interests, ideas, and government preferences, justifying the case selection. Then in section two, I elaborate the blended and focused ideal types. In the third part, I analyze Brazilian positions in the DDA talks and make brief reference to empowerment of the MDIC—an economic-focused bureaucracy—vis-à-vis the Itamaraty in the 2010s, after the analyzed period. The conclusion discusses suggests new avenues of research based on the limitations of those ideal types.

An Empirical Puzzle with Theoretical Implications

Brazil’s offensive demands had as main goal expanding market access abroad for its commercial agriculture, particularly in developed countries, such as EU members and the US. This, however, meant a trade-off at the expense of industry, as the EU and the US demanded the removal of trade barriers in the developing world for their manufactured goods (Gallagher 2007, p. 76). “We knew that, in the round, industry would be sacrificed,” summarized a senior bureaucrat with close connections with the manufacturing sector when interviewed on 6 July 2012. Such a trade-off is puzzling given that, although not as strong as it had been before economic liberalization (Bresser-Pereira 2009), industry remained an important sector for the Brazilian economy, at least as much as the emerging agribusiness segment (Hopewell 2013).

Conventional institutional arguments cannot explain Brazilian preference for liberalization. The MRE and the President (who retains the constitutional prerogative of conducting foreign affairs) considered diverse inputs in formulating foreign trade policy. The Ministry of Agriculture (MAPA) had links with agribusiness, while the Ministry of Development, Industry, and Foreign Trade (MDIC) provided a channel for defensive industrial interests. The supposed “natural” preference of the Brazilian state for defending agricultural exports therefore does not hold: both agricultural and industrial sectors had access to policymakers. Furthermore, the MRE was aligned with the import substitution project (Sikkink 1991), which is inconsistent with the diplomats’ the country’s supposed preference for liberalization in the DDA. Moreover, successive interactions with the WTO system supposedly enhanced Brazil’s capacity to elaborate foreign trade policy (Shaffer et al. 2008). Yet, such was not the case of diplomacy in the DDA given that negotiators conceded too much in industry in exchange for market access in agriculture.

Interests do not provide insights into such a preference either, as both liberalizing and protectionist sectors lobbied the government during the DDA negotiations. Business organizations in industry had systematically been following foreign trade policymaking since mid-1990s, with the negotiations aimed at forming the Free Trade Area of the Americas (FTAA). In 1996, the National Confederation of Industry (CNI) founded the Brazilian Business Coalition (CEB) with the goal of influencing the government in the FTAA talks (Carvalho 2003; Mancuso and Oliveira 2006). The CEB along with the Federation of Industries of the State of São Paulo (FIESP)—where about 40 percent of Brazilian industry is based (SEADE 2012)—aggregated manufacturers’ demands.1 Agricultural commodity exporters adopted the same strategy, with segments (e.g., beef and soya bean producers) being as active as the National Confederation of Agriculture (CNA). However, peasants and small producers took a different position and demanded protection through the National Confederation of Agricultural Workers (CONTAG), which was expected to be as strong as the lobbying structures that represent large farmers.

In Brazil, the preference for favoring agriculture in the DDA is often explained by liberalizing economic ideas. Under such a view, the Brazilian economy had an inherent comparative advantage in agricultural commodities. However, in the first 13 years after liberalization in Brazil (1990–2003), industry’s participation in exports of goods and services remained higher than that of the agricultural sector, contributing at least to 50 percent of all goods sold abroad (World Bank 2010). The Brazilian process of industrialization was one of the most successful industrializing experiences after the Second World War. Industrial output grew four times between 1965 and 1980. Such a rate is even higher than in Japan, which expanded three times in the same period (World Bank 2010). Also, when the DDA was launched, Brazil was moving towards an export-led strategy that included an emphasis on both agricultural commodities and manufactured goods (Rodrigues Vieira 2014), with domestic incentives to both segments (Ban 2013).

Finally, partisan preferences also do not provide a straightforward answer for Brazilian positions. The DDA negotiations started under Fernando Henrique Cardoso’s government (1995–2003), whose Brazilian Social Democracy (PSDB) party was reputed to be more favorable to economic liberalization than his successor, President Luiz Inácio Lula da Silva of the Workers’ Party (PT) (Ban 2013). Despite PT strong links with industrial workers and small peasants, prioritization of gains in agriculture continued under Lula. The lack of effectiveness of such links is puzzling considering that, while capitalists could shift capital from a sector to another in the case manufacturing crumbled, labor would have faced mobility problems. Moreover, under Lula, Brazil faced stronger competition from Chinese manufactures, and agricultural commodities began to dominate Brazilian exports.

If not these arguments based on institutions, organization of interests, ideas, and government preferences, what thus explains Brazil’s national interest in the DDA talks? A hint lies in exploring in further detail institutional factors. Given the centrality of the MRE in defining positions in trade negotiations, its structure and position within the overall architecture of the state may explain why Brazil had offensive interests at the multilateral level. Current institutional arguments, however, face limitations in explaining how state institutions impact the formation of the national interest, which cannot be conceived as stable throughout time. Such limitations, in turn, demand further theoretical developments on how bureaucracies matter in crafting the national interest. With the considerations above in mind, I henceforth build on the institutional literature to elaborate the blended/focused typology and its theoretical implications. In doing so, I first examine arguments that conceive institutions as intervening factors that create incentives for societal actors such as interest groups and NGOs. Thereafter, I elaborate the concepts of blended and focused bureaucracies.

An Alternative Framework: Blended vs. Economic-Focused Bureaucracies

To identify the conditions under which a given set of interests prevails, a starting point is to apply veto player theory to trade negotiations. O’Reilly (2005) concludes that the existence of many veto players reduce the probability of changes in tariffs. Under such account, international factors and the action of interest groups do not explain trade policy. Nevertheless, in focusing on the number of players who can veto a decision, the theory overlooks a plausible scenario in which that numerous veto players could reach a common position if they share interests. The recent argument on access points in state bureaucracies has similar limitations. According to this logic, the more points of access in the state lobbies have, the more likely they are to influence policy outcomes as the costs of lobbying decrease (Ehrlich 2011). Although it is parsimonious, the theory ignores the unequal weights different bureaucracies have. Thus, not all bureaucracies that function as access points for lobbying groups are equal in terms of power. For instance, in the case of Brazil, the MRE is expected to have far more weight than other ministries in crafting negotiating positions. Therefore, although other bureaucratic units that take part in the formation of the national interest, they are neither veto players nor access points through which lobbying could be effective.

A solution for the shortcomings of the veto players and access points theories lies in identifying which sectors of the state are central in international trade negotiations. If a given section of the state has more control over the negotiating agenda than others, such an institution can therefore have more power in defining the national interest. That is the case, for instance, when the executive branch of a democratic government has the prerogative of negotiating international agreements without suffering a constant scrutiny from legislators. According to Lake (1988, p. 57), in the US, the executive finds itself in a position that enables it to decide on foreign policy even against the will of societal actors. Milner (1997) corroborates this argument by arguing the executive branch empowers itself by controlling information related to negotiations.

In any country, however, the executive is hardly a unitary actor, having bureaucracies with different degrees of autonomy. While the head of government can be considered the chief negotiator, other officials deal with negotiations more frequently. Yet, even if the executive prevails, it is not necessarily the only government branch that define the national interest. In democratic regimes legislators may participate not only in the ratification of agreements, but also in its negotiation. The legislative, indeed, plays a role in negotiations even in the case that it abdicates the right of participating in the formation of the national interest rather than formally delegating it to the executive (Martin 2000, p. 148) In addition, certain institutions and actors—such as elected officers in executive and legislative branches—may have the power to define the whole institutional design itself, and, thus, change which section of the state holds the status of the diplomatic bureaucracy.

Therefore, should those institutions and actors be the analytical focus for explaining national interest in trade? The answer is “no” if two assumptions empirically hold. First, both bureaucrats and elected officials can hardly change the existent institutional design: that can be too costly depending on the number of effective veto players and divert political resources from the major goal—negotiating an agreement. In fact, institutional designs are usually inherited from past decisions (Goldstein 1988). Second, if a new institutional design is politically feasible, the shadow of the future might lead policymakers to avoid risk and pursue the desired outcome in negotiations by other means. Institutional changes open room for unintended consequences (Pierson 2000). Thus, a priori there is no need to problematize why a given institutional design exists in a country as long as the same is not modified just because of a specific international negotiation or in function of vested interests.

The discussion above corroborates Miles’ Law (1978): decision-making depends on where one sits in the state apparatus. Such proposition follows Allison’s (1971, p. 171) bureaucratic model of decision-making, in which “each player pulls and hauls with the power at his [or her] discretion for outcomes that will advance his [or her] conception of national, organizational, group, and personal interests.” Thus, rather than being the consequence of careful rational action, the national interest arises from multiple interactions shaped by pre-given factors, such as the state’s institutional design.

Complementing Miles’ claim that “where you stand depends on where you sit,” I propose that how one sits within a bureaucracy also impacts decision-making. If institutions are not easily changeable, thus one cannot choose where and how to sit when deciding. The perceptions of one who decides mixing political-economic issues are different to those of whose focus lies on economic questions only. A focus on the state, in particular the diplomatic bureaucracy, implies not treating its institutional design as an intervening factor. Nor it ignores government branches other than the executive. Instead, the design of domestic institutions has an independent—yet not exclusive—impact on trade positions. That is the case if the existence of the institutional design precedes a given trade negotiation and, thus, the definition of the national interest. I then expect that, insofar as the state formulates foreign policy, it ultimately defines what the national interest means (Allison 1971; Krasner 1978). Yet current accounts leave unanswered why domestic and international constraints impact differently such a process.

Chorev’s (2007) concept of fluid divide attempts to overcome such limitation in conceiving both levels as being part of a single field of action. Notwithstanding its sophistication, the argument differs from the categories I work with. According to Chorev (2007, p. 660), “institutional arrangements of both states and international organizations affect the relative dominance of the competing factors involved in the process of policymaking.” Therefore, the analysis of the institutions involved in trade policy in both domestic and international levels contributes to unfold the logics behind the shifting balance that attributes different weights to national and external constraints. In contrast to Chorev’s argument, however, I contend that an institutional perspective can be more parsimonious if focused only on the domestic bureaucracies devoted to international trade negotiations.

Within such a model, both domestic and international factors are subsumed into a single framework, and it is possible to incorporate constraints that arise from within the bureaucracies themselves independently from societal pressures or foreign limitations. As explained in the previous section, agency shapes institutions, yet the latter evolve in unpredictable ways, opening room for unintended consequences. Such a fact implies that institutional design impacts policymaking independently from other factors as it embeds social action, framing decision-makers’ understanding of the negotiating conjecture. Out of all institutions, the bureaucracy responsible for negotiations is the section of the state that matters the most for understanding the formation of the national interest: it controls the negotiating agenda and, thus, constrains demands coming from other institutions. Other bureaucracies and the legislative power can still impact the formation of the national interest, yet will be subject to the constraints imposed by the design of the diplomatic bureaucracy.

That is the case as the weight of domestic and international factors upon decision-making depends on the institutional design of the state. This, in turn, determines whether domestic or international embeddedness prevails within a bureaucracy. Originally, the concept of embeddedness focused on the autonomy of the state vis-à-vis private interests in the domestic level (Evans 1995). The concept implies that the more embedded a state is, the more sensitive it is to demands from domestic players. For the sake of simplicity, the notion of autonomy can be left aside once the concept of embeddedness is converted into domestic embeddedness and contrasted with international embeddedness. They generate the focused and blended ideal types respectively. While blended bureaucracies are devoted to foreign policy in general take into account in decision-making foreign constraints mainly, focused bureaucracies focus above all domestic interests (including from other parts of the state, such as the legislative power in democracies) in decision-making given their specialization on economic-related issues. Figure 1 schematizes the argument, representing how each type of bureaucracy finds itself in relation to the domestic and international level.

Figure 1: Blended and Economic-Focused Bureaucracies

Certainly, those are ideal types that may not capture all variations in the domestic institutional design of decision-making on international trade negotiations. However, the identification of the diplomatic bureaucracy, followed by the definition of its main characteristic, furnishes basic elements for a rigorous, systematic explanation of the origins of a country’s national interest in trade without ignoring the effects of competing explanations for the formation of the national interest.

Applying the Framework

“Insofar as the MRE faces the limits of foreign positions, it is reasonable that the ministry formulates the national demands in a trade negotiation.” This is how a senior bureaucrat interviewed on 20 June 2012 explained the prominent role the Brazilian Ministry of Foreign Relations played in relation to the DDA talks. As the negotiations evolved, the MRE however had to deal with an increasing number of institutions and non-state actors contributing with policymaking (Cason and Power 2009, p. 118). Therefore, how was it possible for the MRE to mitigate protectionist demands, such as those from manufacturing sectors, in the formation of the national interest? The answer is straightforward: the ministry retained the prerogative of conducting Brazil’s foreign affairs and remained more embedded in the international level than in the domestic arena, defining the national interest based first and foremost based upon the broad international political conjecture. This limited the influence from other bureaucracies and business segments whose interests that could hinder the use of the DDA as a means of politically empowering Brazil in the world stage. In turn, such a project depended on defending liberalizing positions to weaken Global North’s stances on trade.

The MRE is reputed to have more autonomy and internal coherence than other parts of the Brazilian state. While diplomats still claim to have a monopoly in defining the national interest, since mid-1990s the MRE receives more and more inputs from other units within the state (Armijo and Kearney 2008). In addition, given the complexities that arise from globalization, diplomats became more receptive to interactions with organized interests as a mean of avoiding misperceptions in negotiations. Those changes occurred, however, while the ministry preserved its status as the diplomatic bureaucracy for trade negotiations. Given that the MRE perceived the launch of a new multilateral round of trade liberalization as inevitable, the ministry consulted with other bureaucracies and societal organizations in 1999. In July of that year, four months before the Seattle Ministerial Meeting for launching the failed Millennium Round, President Cardoso created the Inter-Ministerial Working Group on International Trade of Goods and Services (GICI). The group focused on liberalization at the multilateral level.2 In 2003, the MRE organized a working group dedicated to agriculture with the aim of building consensual positions for the sector, which was the Brazilian focus in the round (Bureaucrat, interview, 12 March 2012). Called the Informal Technical Group (GTI), it included players other than the MRE (Luiz Carmona, MAPA bureaucrat, interview 19 June 2012). The GTI was built upon the pre-existing connections between the MAPA and market actors (Carlos Cozendey, MRE bureaucrat, interview, 26 June 2012; Camila Sande, CNA Officer, interview, 27 June 2012). The group surpassed the GICI as the main forum for formulating Brazil’s core positions for the DDA negotiations. As diplomats report, many positions that would later be presented at the negotiations by the Agricultural G-20 were first discussed at the GTI (Bureaucrat, interview, 12 March 2012; interview with senior bureaucrat, 18 October 2012). Moreover, the MRE contributed to the creation of a private think-tank called ICONE, which improved the negotiators’ technical background on agriculture (André Nassar, ICONE Officer, interview, 20 July 2012).

A similar approach was taken with regards to industrial sectors. The MRE created a group for Non-Agricultural Market Access (NAMA) themes analogous to the GTI, which focused on agriculture (Market Actor, interview, 4 June 2012). Also, at the beginning of 2004, the Permanent Mission in Geneva pressed the FIESP to expand its research capacity on international negotiations. Diplomats in the Permanent Mission wanted to build critical mass to negotiate NAMA, and, despite initial opposition from top-tier diplomats based in Brasília, the FIESP sent an official to Geneva to work with the Brazilian representatives to the WTO. The role of FIESP as an informal think-tank for negotiations involving industrial goods was crucial insofar as the MDIC opposed the trade-off that the DDA implied for Brazil (Frederico Meira, FIESP officer [2005-2013], interview, 2 August 2011; Market Actor, interview, 18 June 2012). As the round seemed to be reaching its end, in 2008, the Permanent Mission liaised directly with business associations to discuss the impact of the Swiss Formula (Gallagher 2007, p. 74-77) for tariff cuts and the selection of tariff lines for the exception list in NAMA (Senior Bureaucrat, interview, 15 June 2012).

Apart from the MRE, the MDIC—an example of focused bureaucracy—emerged during the DDA timeframe as the most relevant player for the elaboration of positions in foreign trade policy. The MDIC was created in 1999, from the Ministry of Industry and Commerce (MIC), which had historical connections with the industrial sector, including the FIESP and the CNI. Those ties, particularly with the CNI through the CEB, prevailed as the ministry was converted into the MDIC, as part of Cardoso’s strategy, launched in his second term (1999–2003), to boost Brazilian exports and to avoid disruptions in the balance of payments (Rodrigues Vieira 2014, p. 151). MDIC’s ministers between 2003 and 2008 talked directly about the DDA with organized interests and entrepreneurs in manufacturing (Interview with Sérgio Amaral, MDIC’s minister [2001-2002], 25 May 2012; interview with Luiz Fernando Furlan, MDIC’s minister [2003-2006], 2 August 2012; Former Bureaucrat, interview, 14 December 2011). Nevertheless, the proposals and studies were submitted to the MRE, which coordinates de facto the formation of positions in the domestic level (Welber Barral, MDIC’s Secretary of Foreign Trade [2007-2011], interview, 18 June 2012). For instance, archives demonstrate that the MDIC sent to the MRE simulations on the effects of tariff cut proposals. Such an exchange indicates that some informational cooperation existed between both bureaucracies, notwithstanding different views on the negotiations (MDIC 2005 and 2006). The MDIC, however, resisted providing all the information that the MRE requested to define the limits of the inter-sectorial bargains that the country could offer and accept (Frederico Meira, FIESP officer [2005-2013], interview, 2 August 2011; Market Actor, interview, 18 June 2012). The MDIC was also the main access point for services, demanding positions from business associations in the sector. The interaction with them was enhanced in 2005 with the creation of the Secretary of Commerce and Services (Bureaucrat, interview, 5 July 2012). In spite of being mainly associated with manufacturing and services, the MDIC also received liberalizing demands from sectors that were part of the agribusiness chain, in particular food-processing (such as industrialized meat), which falls between the agriculture and industry (Welber Barral, MDIC’s Secretary of Foreign Trade, 2007-2011, interview, 18 June 2012).3

Agribusiness, however, had its main links with the MAPA. As happened between, on the one hand, the heads of the MDIC and, on the other, the industrial sectors, major agricultural producers and organized interests had direct links with the minister of Agriculture. Marcus Vinícius Pratini de Moraes, minister of Agriculture when the round was launched in 2001, reported during an interview on 13 August 2011 that he talked directly with associations of agro-export sectors inputs to formulate initial positions for the round. Given the rising importance of commodity-exporting for Brazil’s foreign trade, the ministry created a Secretary of International Relations to enhance its policymaking capacity in the area (Bureaucrat, interview, 20 June 2012). Small farmers, in turn, had their interests represented by the Ministry of Agrarian Development (MDA). Created in 1999, the ministry only became involved in the DDA negotiations during Lula’s government. This was part of a strategy of listening to the views of the small producers that had defensive interests, thus conferring more legitimacy to the formation of negotiating positions (Bureaucrat, interview, 20 June 2012; Luiz Vicente Facco, CONTAG director, interview, 3 July 2012). The clearest signal of MDA’s incorporation in the process of policymaking in trade negotiations came in 2005, with a Presidential Decree that placed the minister of Agrarian Development in the Council of Ministers of the Chamber of Foreign Trade (CAMEX) (Presidência da República 2005), reflecting PT’s links with peasant organizations and landless movements.

The CAMEX had been created in 1995, just after Cardoso became President. Initially placed within the presidential office with the goal of acting as a coordination forum in foreign trade policy (WTO 2009, p. 15), it was transferred to the MDIC in 2001, as the head of the ministry was given the responsibility of presiding over the chamber. Officially, the CAMEX has the prerogative of determining the negotiation directives for international agreements related to trade issues (Presidência da República 2003). The main decision-making body within the chamber is the Council of Ministers, originally formed by the ministers of Development, Industry and Foreign Trade (the chair); of the Civil House (the President’s chief-of-staff); Foreign Relations; Finance; Planning, Budget and Administration; and Agriculture (Presidência da República 2001). In spite of officially controlling the CAMEX, the MDIC is far from having the last word on trade negotiations. Neither the ministry nor the chamber was part of the GTI that formulated positions in agriculture (Luiz Carmona, MAPA bureaucrat, interview, 19 June 2012). Insofar as the MRE remains in the charge of negotiations (Carlos Cozendey, MRE bureaucrat, interview, 26 June 2012), it retains the coordination position that is legally under CAMEX umbrella. Figure 2 exemplifies the blended model in Brazil, with international embeddedness prevailing due to MRE stronger linkages with external rather than domestic processes.

Figure 2: A Blended Model (Brazil)

Without the existence blended institutional design, Brazilian negotiators would hardly have downplayed the demands of protectionist segments as they were organized as an interest group (i.e., CEB and FIESP) and had connections with relevant sections of the state (i.e., MDIC). Within the international conjecture in which negotiations took place, Brazilian diplomacy considered the DDA an opportunity to enhance the country’s gains not only in what concerns political status as an emerging power, but also in material terms through the expansion of agricultural exports. At the same time, with active participation within the WTO system, the country could signal to the international society commitment to multilateralism. However, under Lula’s government, such a commitment was converted into a means to attempting to enhance leadership among developing countries.

Having been crafted under the perception that the West’s power could be soft-balanced (Hurrell 2006, p. 16; Hurrell and Narlikar 2006, p. 431), the articulation of the Agricultural G-20 satisfied PT’s aspirations to establish deeper relations between Brazil and the Global South (Burges 2009, p. 160-161), MRE left-wing-leaning bureaucrats, and entrepreneurs who had gained strength with the liberalization of the economy. However, that seemed odd under traditional ideas of realpolitik insofar as commodity exporting is hardly associated with state empowerment in the international arena (Gilpin 1981). The FTAA, in turn, which was in the interest of industrial sectors, never came into force in large part thanks for to the Brazilian government negotiating tactic of postponing substantial decisions until the US abandoned the project in 2005. These tactics met part of PT’s anti-American stances, yet came about largely due to MRE’s blended institutional design. As a senior bureaucrat interviewed on 15 June 2012 summarized, “the circumstances allowed ourselves to oppose the EU and the US … In sum, it was a magic moment which allowed us to project ourselves abroad.” Another senior bureaucrat interviewed on 6 July 2012 went further and said that “…the conclusion of the round would enhance Brazil’s diplomatic prestige in the world.”

Developments in the aftermath of the analysed period corroborate the argument that an institutional design centred on a focused bureaucracy would have tamed Brazil’s liberalizing impetus in Doha. Certainly, international factors—particularly the negotiations of the now-defunct Trans-Pacific Partnership (TPP) and Transatlantic Trade and Investment Partnership (TTIP)—contributed to the adjustment of Brazilian priorities in trade. Yet, the rising importance of the MDIC in the issue-area precedes such changes. Under President Dilma Rousseff’s (2011-2016) administration, the MDIC conquered more space in the formulation of the national interest. An example was the pursuit of bilateral trade agreements by Minister Armando Monteiro, who was a former CNI president and commanded MDIC during Rousseff’s second term (January 2015-May 2016, when she left office after the approval of impeachment proceedings). Amid the economic crisis that triggered mass discontent against PT’s just re-elected president, Brazil signed with Uruguay a free-trade agreement to liberalize the commerce of automobiles—a sector subject to a special regime within MERCOSUR, the common market that those countries integrate (Brazil 2015). In the same year, Brazil and Colombia managed to create a quota-based regime for enhancing exchange in the same sector (Ibid.). Moreover, in April 2016, MDIC and Itamaraty signed an economic deal with Peru comprising even government procurement (Brazil 2016).

Those facts signalled that the government finally opened itself to the demands of manufacturers to look beyond MERCOSUR for new exporting markets, yet without concessions at the multilateral level (CNI 2014). Part of these demands steam from the exhaustion of the new developmental policies that stimulated domestic consumption as a means of circumventing the effects of the 2008 crisis on the Brazilian economy (Ban 2013). Those trends only gained further traction within Michel Temer’s government and the emphasis on economic diplomacy chancellor José Serra put during his short-term tenure in the Itamaraty (May 2016-February 2017). Both CAMEX and the Brazilian Agency for Export Promotion are placed under Itamaraty’s umbrella since Temer became president. As of February 2017, however, the MDIC remains active in negotiating bilateral agreements, as its attempts to celebrate a deal with the European Free Trade Area suggest (EFTA 2017).

To summarize, since early 1990’s Brazil witnessed the emergence of partisan preferences on foreign policy as well as of state actors other than the MRE—particularly the MDIC—in dealing with external affairs and the growing interest of organized groups in external economic relations. The MRE, however, retained during the 2000s its dominant position in defining the national interest in trade talks. As the ministry is embedded more internationally than domestically, it looked during that decade mainly at foreign constraints when elaborating the national interest in trade. Developments in 2010s corroborate such an argument as Brazil started to explore bilateral solutions in trade as the MDIC gained more prominence in the negotiation of commercial agreements. With U.S. President Donald Trump’s retreat from trade openness, one could expect that Brazil’s retains the pace of recent change in the institutional design of foreign policymaking and, hence, in its growing bilateral approach in that issue-area. The attempts of approximation between MERCOSUR and the Pacific Alliance—which congregates Chile, Colombia, Mexico, and Peru, all of which Brazil already has deals with—suggest such a prediction is plausible.

Conclusion

In this article, I argued that the design of the bureaucracy at the forefront of trade negotiations decisively impacts the formation of the national interest in such issue-area. As bureaucrats craft the national interest, they are subject to constraints that have different weights on decision-making. Whether international or domestic factors prevail depends on the configuration of the diplomatic bureaucracy, which ultimately gathers domestic demands and represents the state abroad. If such a bureaucracy oversees foreign policy in general, international embeddedness prevails over domestic embeddedness, thus leading strategic considerations to prevail over immediate market demands in policymaking. In turn, the opposite happens whenever the main bureaucracy is focused on economic issues—leading a country to balance better distinct interests from liberalizing and protectionist segments. As the institutional design precedes the preferences of organized interests, bureaucrats, and politicians, such typology explains the formation of different notions of national interest.

The argument was built upon Brazil’s case in the negotiations of WTO’s DDA. The MRE conducts trade negotiations, representing a case of focused bureaucracy, then being more subject to international rather than to domestic forces. Brazil therefore viewed the DDA negotiations as a means of enhancing Brazil’s prestige abroad rather than prioritizing the demands of protectionist segments, not offering second-best options for liberalizing segments such as commodity-producers. As the DDA went into deadlock, domestic factors gained more influence once the MDIC—an economic-focused bureaucracy—participated more in trade negotiations, thus suggesting that, had it controlled the negotiating agenda during the DDA, Brazil could have ended up balancing demands from sectors with opposed interests, creating more space for the demands of manufacturers.

Three main research avenues can be explored based on the ideal-types I developed. The first consists of exploring cross-temporal interactions inside states, particularly between executive and the legislative institutions, as the national interest is formed. In Brazil, no evidence on the impact of the legislative in forming the country’s positions in the DDA arose from the interviews and archival research conducted for this work. Based on the US case, however, one could expect that legislative control over decision-making would mitigate the effects of a blended design as representatives would pressure diplomats to prioritize economic needs of constituencies over strategic goals in the light of factors other than market gains. The US legislative often delegates to the President the power of negotiating agreements through the Trade Promotion Authority (TPA). Yet, rather than leading the legislative to abdicate to its role in foreign policy (Martin 2000) and granting full autonomy to the President—who then appoints a Trade Representative (USTR) to negotiate on the government’s behalf—, TPA makes the executive more accountable vis-à-vis legislators (O’Halloran 1993; WTO 2008). In fact, during the TPA mandate that lasted between 2002 and 2007 the USTR started defending liberalizing positions, yet the American space for bargaining shrunk as domestic actors became more protectionist.

A second avenue corresponds to applying the blended/focused typology to other cases in issue-areas other than trade. In the case of Brazil, the effects of blended diplomacy are also evident in the approach adopted during the PT government—particularly under Lula—towards Latin America and Africa and in the process of creating of national champions in agriculture and services. The pursuit of partnerships with neighbors and the other side of the South Atlantic aimed to strengthen Brazil’s position vis-à-vis both established and emerging powers, yet co-opted private actors, such as civil construction subcontractors (Rego and Figueira 2017). The strategy also relied on state owned companies, particularly the oil-giant Petrobras, as the disputes over gas fields in Bolivia (Ribeiro 2009) and the exploration of off-shore reserves in West Africa suggest. Moreover, in current times the opposition to Brazil’s application to become a full member of the Organization for Economic Cooperation and Development (OECD) can be understood within the context of blended diplomacy the MRE has historically conducted. Those against the application usually refer to the trade-offs between potential economic gains that OECD may bring and an eventual loss of autonomy in conducting key areas of foreign policy, particularly cooperation and development aid—for which that organization demands more transparency than what Brazil practices. Should Brazil have a diplomatic unit focused on economic affairs, membership in the OECD could perhaps have not been perceived under such a zero-sum game logic.

The third research line the blended vs. focused typology posits consists of exploring the impact of different types of international institutions on the formation of the national interest. The case I explored comprises an issue-area (trade) and an organization (WTO) with a high degree of legalization (Abbott et al 2000). This implies that the state does not have as much flexibility as it would have in the case of negotiating an agreement from scratch or even the formation of a new organization. Such a context may lead diplomacy to control the agenda of the formation of the national interest to extract as many gains as possible from negotiations not only in strategic terms, but also for what negotiators perceive to be the best to domestic actors. However, the Brazilian experience under the GATT—which was not as institutionalized the WTO—suggests that the domestic institutional design trumps any potential influence from the design of international organizations in the formation of the national interest. As mentioned in the introduction, Brazil have always bargained hard with the West in the multilateral system of trade. India’s case corroborates such an argument as it has a similar trajectory in the GATT/WTO system, having been, however, much less flexible than Brazil, standing with protectionist positions even whenever the country could have strategically embraced liberalization (Rodrigues Vieira 2015). Indeed, such a fact may reflect the institutional design of the Indian diplomatic bureaucracy for trade negotiations, which is economic-focused as the Ministry of Commerce negotiates trade agreements for the country and interacts constantly with other bureaucracies and interest groups.

International phenomena are subject to multiple influences, being hardly the outcome of a chain of deliberate rational decisions only. Domestic institutions, thus, have an independent effect on the process of crafting national preferences and in absorbing external influence. As institutional design does not necessarily reflect the choices of those who participate in policymaking, but a previous political settlement, bureaucrats and politicians usually have no better choice than conforming themselves to the environment in which they are embedded. The same design determines whether bureaucrats will be primarily subject to international or domestic factors as they define the national interest in trade negotiations and eventually the overall strategy of empowerment in the international arena.