Debunking Brazilian Exceptionalism in its Africa Relations: Evidence from Angola and Tanzania

ABSTRACT From 2003, President Lula heralded a new dawn in Brazil’s expanding African relations. Brazil was claimed to be unlike other exploitative powers because of its cultural, geographic and historic connections; Africa’s true brother. Despite the passing of two decades and a number of scandals, this narrative of exceptionalism remains. Studies on Brazil–Africa relations tend to focus on the Brazilian state as the key, essentially benign agent. Our analysis uses the case studies of Angola and Tanzania to debunk the idea of Brazilian exceptionalism. We demonstrate the significant, overlooked agency of corporations in shaping and implementing Lula’s Africa Policy, and determining its developmentally dubious outcomes. Additionally, the paper shows how political elites in Africa directed Brazilian government and companies into their political and business norms. Thus, Brazil–Africa relations replicated much of the typical economic patterns of the continent’s trade, with oft-controversial and corrupt investment in commodity extraction and infrastructure.


Introduction
Over the last two decades, Brazil's engagement in Africa has swung markedly. Trade rose from $4.2billion in 2000 to a peak of $27.6billion in 2011 (Stolte 2015). Meanwhile, diplomatic links exploded. During Luiz Inacio Lula da Silva (Partido dos Trabalhadores/ Worker's Party [PT]) 2003-2010 presidency, 520 official government-to-government agreements were signed 1 whilst embassies rose from 17 in 2005 to 37 in 2010 (Stolte 2012(Stolte , 2015Alves 2013). Brazil also became a leading South-South Development Cooperation partner (Carmody 2011), with 60% of Brazil's Development Cooperation Agency's (ABC) spending allocated to Africa. 2 Former Brazilian President Luiz Inácio "Lula" da Silva rhetorically-elevated the importance of these ties, for instance proclaiming in 2006 that "Today, Africa is an indisputable priority". 3 However, this zenith proved short-lived. The rapid acceleration collapsed from 2014 under the shadow of Brazil's economic recession (Stolte 2015, 159) and political tumult initiated by Operação Lava Jato, a major corruption investigation implicating much of the political-business elite. Although now seemingly consigned to Brazil's past, detailed accounts of this period remain sparse. Those that do exist tend to emphasise the pre-eminence of the Brazilian state in forging this trajectory. To some extent this perpetuates narratives asserted by the Worker's Party (PT) that Brazil-Africa relations were exceptional: They would not replicate Africa's past international relations which tended to create structural dependency and exploitation for the benefit of elites. Rather, Brazil presented itself as an equal, a true brother to Africa enabling both Brazil and the continent's economic rise (Hurrell 2013). This rhetoric of South-South Cooperation is not unique to Brazil, being rooted in the Third World, Non-Aligned Movement that formally began in 1955 and pledges to uphold mutual support between developing countries (Carmody 2011;Mawdsley 2012). President Lula and other Brazilian government officials took this rhetoric a step further. They claimed to have unparalleled knowledge of Africa given certain climatic similarities and the influence of the continent, through the slave trade, on Brazil's contemporary cultural milieu. In contrast to business-orientated, extractive-focused world powers such as China and America, Brazil would export its "Brazilian dream", its set of economic and socially-progressive policies that increased growth whilst reducing poverty, hunger and health issues like HIV/AIDS.
This article critiques such assertions of exceptionalism, showing the difference between rhetoric and practice. Using the broader literature on Africa's engagement with the international system as a theoretical basis, we make three key contributions. First, we suggest that Brazil's ties, whilst having their own individuality, represent continuity with, rather than exception from, the continent's norm. Brazil has reinforced Africa's structural economic position in the global economy, frequently characterised as one of extraversion, whereby countries on the continent tend to exchange primary commodities for manufactured goods and services. Second, rather than privileging Brazilian actors, as in much of the literature on Brazil-Africa (e.g. Stolte 2015; Alves 2013; Freitas Barbosa, Narciso, and Biancalana 2009;Garcia and Kato 2015), we demonstrate the strong influence of governments in Africa, drawing on the literature stressing the role of African agency in the continent's international relations, even when in positions of apparent weakness. Third, building on others who have sought to differentiate Brazil as a homogenous player in Africa (Milhorance 2019; Nogueira and Ollinaho 2013), we assert the importance of the country's companies in pushing forward relations, especially those in the construction and extractive sectors, demonstrating the degree to which state and private-sector actors joined forces to build international engagement. We therefore use mainstream African-studies analysis of the continent's international relations as a theoretical lens through which to view Brazil-Africa ties, demonstrating the degree to which Brazil fitted into Africa's typical patterns. This argument also evidences the way in which Brazil exported its state-business model to Africa. Particularly from Lula's second term (starting in 2006) a corporate turn occurred whereby the PT government closely engaged with large mining, infrastructure and agriculture firms to ensure domestic political survival and the implementation of their policy vision. This argument is demonstrated by case studies of Angola and Tanzania, chosen because they capture the range of Brazil's new 21st century ties. Angola is Brazil's longest-lasting African partner, with post-colonial ties starting at the country's 1975 independence. In Angola, state power, especially after President José Eduardo Dos Santos' accession to the Presidency in 1979, increasingly flowed away from formal channels to the Angolan presidency. 4 Dos Santos and his close entouragereferred to by Angolans as the Futungo after a presidential compound in the outskirts of Luandaformed the country's political and business elite. Certain Brazilian actors were well-integrated into this group, especially construction firm Odebrecht, who were tasked with early investment in sensitive areas during and after Angola's 1975-2002civil war. Indeed, by 2002 the company was already the biggest donor of the Eduardo dos Santos Foundation, a supposedly philanthropic organisation that served as a revenue generator for the Angolan elite and it was this same institution that hosted Lula after his departure from office. 5 In contrast to this historic underpinning and the shared-Portuguese language, Tanzania represented an essentially new engagement for Brazil, as it had little experience or knowledge of the country. Until 2017, its politics also contrasted with Angola as although dominated by one-party, CCM, significant fractures between rival politic-business groups in the country's ruling elite created a different political and policymaking reality. Therefore, the two case studies enabled this article's analysis to capture important dimensions of the breadth of Brazil-Africa relations.
This article is based on extensive primary research. It utilized elite interviews, defined as conversations with informants who were chosen for the position they occupy and knowledge of internal politics. In Angola, research was undertaken between 2009 and 2012 in Luanda and the Lunda diamond-mining provinces with businessman and party cadres that gravitated around the family of former president dos Santos (Maussion 2019). In Tanzania fieldwork between 2015 and 2016 involved over 60 semi-structured interviews in Dar-es-Salaam with both senior officials (directors, former ministers and CEOs) in government ministries and agencies and with Brazilian companies and diplomats. In addition, both authors have carried out over 53 interviews in Rio de Janeiro, Sao Paulo and Brasilia with senior diplomats, directors in the Federal Ministries, researchers, NGOs and company officials between 2012 and 2019. This article has also benefited from insights from one author's position as senior advisor to the secretary of strategic affairs during the second Dilma Rousseff government (2015)(2016), which provided direct access to those civil servants carrying out the Africa strategy. To triangulate their findings, we reviewed grey literature and news reports. Conclusions were built using an inductive, theory-building approach 6 that used "Africanist" literature on the continent's global position. We found interviews with government officials and private sector employees particularly illuminating as they demonstrated the bargaining power of African elites in negotiations with Brazilian actors, orientating foreign investment towards their own political priorities, as well as the ability of Brazilian business to influence decisions at the top of government.

Africa, the emerging powers and the international system
Not unjustifiably, texts on Brazil-Africa (Stolte 2015;Alves 2013;Freitas Barbosa, Narciso, and Biancalana 2009; have frequently stressed the role of 4 The internal evolution of the MPLA regime is beyond the remit of this thesis but see, for instance, the work of Messiant (2001) and Soares de Oliveira (2015). 5 The event hosted by Odebrecht itself. 6 Comparatively detailed by Gerring (2017).
Lula and the PT. In contrast, a longstanding strand of Africanist literature has sought to dispel myths of the continent as helpless recipient. For instance, researching the World Bank and IMF's 1980s structural adjustment programmes, Mosley, Harrigan, and Toye's (1995) classic text demonstrates the power of governments in Africa to shape negotiations, even when heavily indebted. Whitfield and Fraser (2009) have similar conclusions about aid negotiations, showing the significant agency governments in Africa have to shape the terms of international engagement, even if not all utilise this power. This literature informs this article's conceptual approach. Even if a relationship appears exploitative for citizens and wider societal interests, national elites can benefit. The concept of extraversion (Bayart and Ellis 2000;Cooper 1981) is one commonly-used way of explaining this. It refers to a deliberative strategy of international economic dependency where a state's wealth, and its political elite's power, is almost exclusively garnered through exports. This dis-incentivises the creation of a domestically resilient economy with its own manufacturing base. Rather than being the defenceless, exploited continent, this approach re-balances understandings of agency in favour of African elites.
We use these analytical insights to understand Brazilian business relations in-Africa. Particularly in more centralised authoritarian states with less transparent legal frameworks and in which the rule of law can be overridden, personal relations with key members of government are vital. Overall, this dynamic has contributed to the domination of primary commodities in Africa's exports (Taylor 2014;Cooper 1981). Such trends have become an important debating point in the literature on China, India and Brazil in Africa. Initially, discussion of these countries' increasing role tended towards doomladen pronouncements about exploitation. Commentaries frequently paint China as a ravenous and indiscriminate force plundering Africa's resources (Power, Mohan, and Tan-Mullins 2012, 6). Numerous scholars (Mawdsley 2012;Power, Mohan, and Tan-Mullins 2012;Taylor 2014;Mohan 2015) then sought to counter these claims, partly through more empirically-evidenced research, but also by showing the hypocrisy of singling out the so-called emerging powers when western companies undertook the same activities. Such claims have also been countered by the assertion of African agency (Mohan 2015;Mohan and Power 2008;Alden, Large, and de Oliveira 2008).
Alongside this progress, the literature on the emerging powers in Africa remains more focused on governmental, not corporate, actors. In the China-Africa case, this partly arises from the prominence of state-owned corporations whose heads effectively rank as cabinetlevel positions in the communist party. Similarly, much of the India-Africa literature stresses the state's role in initiating foreign policy, whilst recognising the importance of the diaspora's independent activity in Africa (e.g. Dubey and Biswas 2016; Chaturvedi and Mulakala 2016). In contrast, Taylor, van der Merwe, and Dodd (2016) assert that large Indian firms, from Tata to Vedanta and Bharti Airtel, dominate India's engagement in Africa. As Carmody (2011) evidences, such strong corporate agency is also typical of European and American engagements; the 2000s' commodity price boom attracted BHP Billiton and Rio Tinto, Exxon Mobil and Total to Africa. Indeed, British-based firm BG was the leader of Tanzania's offshore gas discoveries mentioned below. In contrast to emerging-power companies, analyses do not link these firms to nation states and yet they frequently shape Western foreign policy. Indeed, the last decade has seen "Western" companies' growing role in aid. According to Mawdsley (2018), this is part of a "southernisation" of aid, where development spending supports domestic corporate interest along the same modalities as South-South engagements. This is evident in the UK's Department for International Development which has increasingly enrolled the private sector in the delivery of development (Mawdsley 2015). However, the literature on Brazil-Africa has largely overlooked companies' roles. With greater attention to party politics in Brazil and African countries, this article shows the importance of this oversight, demonstrating the extent to which Brazilian companies forged the country's Africa policy and undertook its implementation. Using this literature as a theoretical lens, the article debunks ideas of Brazil in Africa as exceptional, demonstrating how it fits a typical structural pattern of international engagement.

Brazil's assertion of exceptionalism and its corporate turn
Existing literature covers many important features of Brazil-Africa relations. Andres Malamud (2011), for example, frames the global significance of Africa ties for Brazil's geopolitics and Oliver Stuenkel (2015) additionally assesses Africa's role in consolidating the BRICS group on the international stage. Marcondes, Kim, and Mawdsley (2017), meanwhile, show the degree of Presidentialism in Brazil's international relations and, alongside a number of other texts (e.g. Milhorance 2019; Cabral and Leite 2015;Cabral 2016;Nogueira and Ollinaho 2013), analyse the rhetoric of South-South-Cooperation and its often-problematic translation into projects on the ground. This article builds on these scholars' insights into rhetoric versus practise and their attention to the different actors that make-up Brazil-Africa relations. However, it makes a significant advancement in paying closer attention to Brazil's elite-level politics, showing the degree to which companies were plugged into the highest levels of government, shaping and implementing the PT's evolving Africa policy.

Optimism, solidarity and brotherhood in the early years
The PT's Africa policy was built on limited, yet sustained, Brazil-Africa engagement (Stolte 2015). In the 1970s, military regimes recognised a number of African liberation movements and supported Brazilian corporations' internationalisation (de Freitas Barbosa, Narciso, and Biancalana 2009), including through government-owned banks like Banco de Brazil (Garcia and Kato 2015). Attention from the state, and the private sector, lessened during the 1980s and 1990s democratic governments, with a number of embassies closed and a diplomatic focus centred on the USA (Seibert 2019a). However, corporate activities continued whilst South-South Cooperation increased under the auspices of the ABC development agency, inaugurated in 1987. During this period, relations with Africa were also secondary for the Worker's Party. Since its creation in the 1980s, the party had generally been aligned with mainstream social-democratic positions on African politics. Party documents from that period show a consistent discursive focus on defending liberation movements and, after the rise of post-Apartheid South Africa, on the promotion of democracy and social justice. 7 Indeed, the Brazilian left often rhetorically linked the ANC and the PT's rise to power as well as the leaderships of Nelson Mandela and Lula. 8 The PT's political contacts in Africa were maintained mainly via individuals including political exiles or via its union, the Central Única dos Trabalhadores, which kept an active partnership with the Congress of South African Trade Unions. Despite this strong rhetoric, the PT did not have a specific Africa agenda when it started governing in 2003.
However, parts of the political coalition behind this electoral victory did have an African interest. Social movements, NGOs and leftist academics started to gravitate around the Worker's Party in the 1980s before becoming fully aligned with its political project in the 1990s, in a process that marked the beginning of the PTs transformation from a militant to a bureaucratic party in the 2000s (Secco 2015). These groups stressed the moral imperative of building Africa ties given the continent's poorer status and given connections through slavery (Gala 2007;Seibert 2019a). Other rationales for greater Africa relations stemmed from cultural links. Atlantic Africa strongly influences Brazilian food, music, dance and art, particularly in the country's north. Additionally, religious links were growing. In contrast to the Catholic Church Brazilian neo-Pentecostal churches such as the Igreja Universal do Reino de Deus promoted themselves across the South and particularly in Lusophone Mozambique and Angola .
Allied to these ideas was an intellectual foreign policy position known as globalismo (Seibert 2019a). Rooted in Non-Aligned Movement (NAM) ideas and popular amongst leftist academics. It asserted that Brazil could better serve its international interests on the world stage through alliances with other developing countries (Saraiva 2011). As one senior official explained, whilst prior governments had "explicitly formulated and defended (the) … [idea that the] path to autonomy was through linking [Brazil] to major political powers [i.e. European and North American countries]", the PT rejected "defining Brazil's position in the world in relation to the dominant centre". Instead, it argued that the country's international political "niche" would stem from South-South relations. 9 This argument was strengthened by the rise of the emerging powers like China, India and Korea over the 2000s, and their increasing influence on international relations. 10 Thus, inheriting Brazil's low-level of engagement, Lula and key PT officials sought to appropriate and redefine Brazil's Africa policy.
Such interests shaped the government's initial guiding principles. In one of his first interviews the Foreign Affairs Minister, Celso Amorim, announced that "Brazil would rediscover Africa" (Rangel 2003). PT politicians followed suit, seeing a turn towards Africa as an extension of their key domestic policy of integrating the Afro-Brazilian population. 11 Moreover, Bolsa Familia, the signature policy for such integration, was claimed to be a "technology that could be exported to Africa". 12 Brazilian officials also asserted the narrative of "Southern" countries as equals, as victims of the powerful north and therefore uniquely placed to offer mutual support for one another. In such rhetoric, Brazil's duty was to be the "brother" to Africa in a win-win relationship (Renzio et al. 2014); a non-interventionist, "champion of the south". 13 As a senior diplomat explained, Brazil preferred engagement and openness to interventionist policies: We "have a view that our presence is better than having an isolated country"; "It is not our way of doing things, isolation, we think the best thing is to promote dialogue". 14 This rhetoric fits with global ideas of South-South Cooperation, but Brazilian officials additionally asserted their exceptionalism in contrast to other countries, including the emerging powers, they presented themselves as supposedly benign and non-exploitative. As one senior PT official stated "the fact is we opened ourselves to Africa [and] (this is) … what is different to other countries". 15 Another added that Brazil operated in contrast to Westerners, who "arrive and act in fundamentally extractive ways"; they didn't have developmental aims. 16 Lula asserted this publicly on a number of occasions. On a 2010 trip in Tanzania, for instance, he claimed that China brings less benefits, for example through employment, than Brazil (Uchôa 2010a). Underpinning these statements was a "common destiny" thesis based on the conviction that Brazil had the answers for Africa, that the PT's socially inclusive project was not only compatible, but also most appropriate for Africa because of shared conditions. Having been joined in Gondwanaland, Brazil has comparable plant species to West Africa and its north contains Savannah zones with a similar climate to parts of Africa. This is particularly illustrated by ProSavana, a "mega-program" carried out by the Japanese international Cooperation Agency (JICA), the Brazilian Cooperation Agency (ABC) and the Mozambican Ministry of Agriculture (MINAG). It attempted to replicate the Cerrado agricultural and development model in Mozambique's northern Nacala Corridor (Cabral and Leite 2015, 435-445). Brazil also has deep cultural connections with Africa through the Atlantic slave trade. Interviewed officials used this to assert that Brazil's northwhere the majority of descendants from Africa live, was comparable to Africa whether in terms of agriculture or the nature of poverty.
The PT therefore presented engagement with Africa as the logical conclusion of its progressive social agenda and reconciliation with Brazil's history. As reflected in numerous interviews with government officials, 17 researchers 18 and company officials, 19 Brazilian experiences of poverty and medical issues at home bred confidence in their ability to diagnose Africa's development solutions: the continent was a known unknown. As Foreign Minister Celso Amorim used to say, "for every African problem there is a Brazilian solution" (Dye 2018, 131). Similarly, a metaphor attributed to Lula was that Africa was thirsty for Brazil, "like rain falling on dry ground … [you can see] how quickly it gets absorbed". 20 Ultimately, by linking Brazil's history and environment to Africa, the PT government aimed to unify its domestic and foreign policy. Extending the same policies domestically and internationally would serve progressive social goals but would also advance Brazil on the global stage, securing its destined great power status as a champion of the Global South. With such unification, Lula's ritual inauguration of grandiose projects across Angola, Mozambique and Ghana not only served to justify his Africa policy, but also underline the success of his national project. Given these assumptions, it is arguably unsurprising that we find little effort to understand local contexts in Africa in the export of Brazilian policies.

2006: the corporate turn
However, this wasn't merely the responsibility of government. The PT and their associated social movements founded the Africa policy with lofty progressive ideas but, with the agency of Brazilian companies and African governments, the policy radically shifted overtime resulting in closer replication of Africa's typical international engagements. A key trigger for this change was the 2005 "Mensalão scandal", involving the revelation that PT paid members of congress to approve legislation. After this, to keep sufficient support in the perennially-fragmented legislature, Lula's government became more pragmatic, forming governing coalitions with more conservative parties 21 and building close ties with big business. This represented a departure, as previously, the PT resisted such firms' campaign contributions. However, as Garcia (2008) shows, corporate financing for Lula's re-election campaign exploded between 2002and 2006. Indeed, according to Filgueiras and Gonçalves (2007 "the largest financiers of Lula's campaign have a leading role in the economy […] the companies of the primary sector are commanding the global insertion of the country via the commodity market". Thus, PT politicians and businessmen built an alliance whereby corporate "donations" ensuring the party's political survival were exchanged for concessional government lending. Africa became the ideal ground for building this new coalition, causing the PT's loosely-affiliated intellectual activist and social movement backers to be replaced by major corporations as key influencers of Brazil's Africa policy. Consistently rising budgets tempered internal governmental contestation over this transition, as did the way companies increased the capacities of development agencies and diplomats to implement the grand Africa vision. Central here were two firms, Odebrecht and Vale. As the drive for commodities reached its peak, mineral-rich, infrastructure-constructing African countries were important markets for both firms. In exchange for their support for his domestic agenda and campaign finance, Lula put his political weight behind business deals for both firms, notably including Vale's bid for coalmines in the latter's Tete province, beating Chinese competition (Rossi 2015). Such close relations increasingly rendered PT's evolving Africa policy an informal public-private partnership; left-wing Lula built an intimate relationship with Roger Agnelli and Emilio Odebrecht, the billionaire CEOs of Vale and Odebrecht respectively.
As well as campaign donations and Lula's lobbying activity, which continued after his Presidency, the President's speeches evidence this corporate turn. Whilst continuing to assert South-South Cooperation, rhetoric shifted to accommodate a greater role for Brazilian business. Whereas initial speeches closely referenced Brazil's black movement and ideas about how to empower Africa, by the end of the 2000s, they increasingly focused on benefits to Brazilian business. 22 One researcher who interviewed him in 2015 remarked that he "didn't want to talk about solidarity" but rather said "he was happy when companies succeed (in Africa)" as they "should replace European countries" because "Brazilian

21
Notably with the PMDB. 22 Interview with Senior Itamaraty Diplomat, 2016. companies could put in place the solidarity idea"; they were more "ethical". 23 Government therefore sought to justify the corporate turn through narratives of Brazil's exceptionalism. As an interviewed senior diplomat claimed, Brazil was unlike China which "has its own way of working … [where] they have little local employment". He stated that one African President reported that "we like the Brazilians because they don't bring the bread, they teach us how to make it". 24 Similarly, an interviewed corporate official stated that "Brazilian companies are a different level (to Chinese) … [there are] some principals [they] don't like to break", indicating that they would only build good quality infrastructure in contrast to others. Thus, Brazil's exceptional status was rhetorically extended to the country's companies.
To cement this corporate shift, the government increased finance for African engagement, something particularly important for infrastructure projects as they typically require concessional credit. Initially, Banco de Brasil provided this but, as regulations were limiting its volume of international lending, the PT government turned to Brazil's Development Bank, BNDES (Capi and Viana 2016;Corkin, Viana, and Naidin 2012). Historically, BNDES had largely shunned this international role given their narrow remit of supporting domestic development (Viana 2016). However, the PT managed to persuade BNDES's then-head, Luciano Coutinho, that it was needed to overcome unfair global competition. As Miguel Peres, the head of Odebrecht's Mozambique operations, explained: "when the topic is the access to credit for exports and services, it is as though the Chinese played in the first league, while us [Brazilians are] in the second" (Goés 2014). Agreeing with this international threat, BNDES changed its financing rules in 2005. Consequently, and alongside a dramatic domestic expansion, international lending for construction companies grew 1185% in ten years. 25 This level of state support continued under Lula's PT successor, President Dilma (2010-2016), who slightly shifted Brazil's Africa Policy; something Seibert and Visentini (2019) call "continuity without priority". In brief, Dilma reduced resources for South-South Cooperation projects, which dropped by half in her first three years, and did not engage in Lula's level of Presidential diplomacy. Conversely, the level of corporate support, for instance through the CAMEX, increased (Seibert 2019b). Dilma's tenure thus consolidated the corporate character of Brazil-Africa relations. Thus, a foreign policy drive that started with principles about reversing, or at least attenuating, the extractive character of Brazil-Africa relations had profoundly changed. However, literature has largely overlooked this, instead of stressing the overwhelming importance of the Brazilian government, and particularly Lula, as the dominant Africa policy driver (Alves 2013;Stolte 2012Stolte , 2015. As the above narrative demonstrates, the PT's social base and established diplomats like Celso Amorim were indeed key to initiating a state-led push to the continent. However, this section argues for an additional appreciation of the agency of companies. As exposed by the Lava Jato and FBI police investigations, Brazil's corporations became increasingly entwined in policy making at home and abroad as the PT's tenure wore on. Such a strong corporate role is not new in Brazilian foreign policy. Private firms, and Odebrecht and Vale in particular, were entwined in successive Brazilian governments, and thereby, their foreign policy (Garcia and Kato 2015). Such relations were established in the 1970s when the Brazilian military regime, faced with the need to increase exports and spur economic growth, built a close relationship with a set of companies styled as the "national champions" and provided support with government contracts, subsidised finance and significant backing for international trade (Campos 2014;Musacchio and Lazzarini 2014). Thus, the change that was cemented by 2006, perhaps best signified in the aforementioned 1185% increase in BNDES lending to construction firms, was in the relationship between the PT and a handful of large Brazilian companies. The influence of leftist-"globalismo"internationalists, social movements and Afro-Brazilian groups within the PT apparatus initiated the first Africa policy grounded on principles of solidarity and mutually-beneficial South-South Cooperation. However, such groups influence waned as, in a bid to shore-up domestic political power, the PT responded to ongoing demands from Odebrecht, Vale and others including large agribusinesses (Milhorance 2019) to be involved in the formulation and implementation of relations with Africaand indeed economic policy more widely. As foreign minister Celso Amorim stated in 2010, "business is business" increasingly defined the Africa policy. 26 Therefore, although marking an important change within the PT, the corporate turn, in many ways represented a return to the norm. The article now turns to the cases of Angola and Tanzania to demonstrate the way Brazil's engagement replicated these countries' typical international relations.

Demonstrating corporate and African agency in Angola and Tanzania
4.1. Angola: the myth of "Odebrecht's empire" 4.1.1. Brazilian agency Despite strong cultural and historic ties, oil drove Brazilin President Ernesto Geisel's first engagement with Agostinho Neto's MPLA in independent Angola, then an internationally isolated pariah embroiled in a longstanding civil war. 27 As a strategic ally of the Brazilian regime, Odebrecht was seen as the most capable company with the technical resources and managerial skills to succeed in unstable Angola. Thus, Petrobrasthe state-owned oil company-created an account linked to Angolan oil in 1985 to pay for exports for the Capanda Dam. The project was Brazil's first "oil for infrastructure" loan and this modality became the preferred mechanism for Brazil-Angola cooperation (Alencastro 2019). Throughout the 1990s, Odebrecht, whose influence in Brasilia survived the transition to democracy, unremittingly lobbied Brazilian presidents to visit Angola. Even those without a specific African policy, such as President Fernando Collor, went to Angola's provinces to express support for Odebrecht. 28 This helped increase diplomatic ties, which expanded particularly under President Fernando Henrique Cardoso (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003). Even with such prominence, Lula brought a significant step change in relations. Supported by the global commodity boom and a series of oil field discoveries in the South Atlantic, Angola's strategic value grew. 29  over his two terms and asserted that Brazil could provide Angola with a socio-economic model. Prominent, political, cultural and media figures close to Lula frequently engaged with Angola, helping to create a narrative in Brazil of Angola, a highly illiberal regime (Soares de Oliveira 2015), as a democracy in-waiting, and President Dos Santos as a man of peace: The first time Dos Santos talked publicly about leaving office was in an interview with Franklin Martins (2013), Lula's former communications official. However, the most substantive impact came from the extension of government credit to Brazilian companies. Despite scarce information (after the Minister of Development, Fernando Pimentel, declared BNDES' Angola operations confidential in 2012) under Lula, 437 of 571 BNDES loans for service exports went to Odebrecht; US$7.7.9 billion of US$12.6 billion (Fernandes and Neder 2019). Lending was particularly directed at Angola, which took 20% of this export portfolio between 2007 and 2014, making it BNDES' largest international country recipient. In August 2019, a major investigation by the federal audit office showed that BNDES had approved Angolan financing in an almost automatic way, with little regard for detail. 30

The symbiosis between Odebrecht and the Angolan ruling elite
Rather than seeing this diplomatic activity and government lending as solely PT's initiative, we assert that in Angola, Odebrecht and the MPLA led Brazilian relations. Starting with the Capanda Dam, Odebrecht earned the MPLA's respect during Angola's civil war. Then in the 1990s, Odebrecht supported the military take-over of Lunda Sul's Kimberlite diamond mine by helping establish the Catoca consortium to run it. Outside the oil industry, Catoca became the most economically-valuable project in Angola and aided the defeat of UNITA (the MPLA's enemy) who had relied on diamond wealth. Odebrecht's special relationship allowed it to join the empresários de confiança, a secretive group mostly comprised of oil companies who influenced key economic policies (Messiant 2001, 287-309). Moreover, the company served as the founding donor for the Eduardo dos Santos Foundation, a revenue generator for the ruling MPLA masked as a philanthropic institution. In a 2016 plea deal former Odebrecht CEO Emilio Odebrecht provided unpresented descriptions of this relationship, relating that by the end of the civil war in 2002, Dos Santos asked Odebrecht to help "convert generals into entrepreneurs"to make sure that members of the military apparatus could take part in Odebrecht's business ventures so that they would "not disturb political stability". 31 Indeed, most of the company's Angolan business consisted of joint ventures with companies that, short of financial resources and technical expertise, functioned mainly as vehicles to channel rents to elite MPLA members tied to the Dos Santos presidency. 32 Moreover, Emilio Odebrecht stated that in 2009 when oil exports collapsed given the global financial crisis, Dos Santos reached out to the Odebrecht family to obtain a Brazilian government line of credit. Whilst these plea deals should not be taken at face value, its narrative corroborates other evidence here to underline the status of Odebrecht in Angola.

30
Federal Court of Accounts, report number 017.469/2016-3, published on 28 August 2019 (https://portal.tcu.gov.br/data/ files/FD/F0/8B/68/02EDC6103FD7BCC6F18818A8/017.469-2016-3%20-%20ASC%20-%20BNDES%20paises% 20estrangeiros%20obras.pdf). This report also shows BNDES commonly overpaid. 31 Brazil's Federal Prosecution Service, file number: 52614/2017-GTLJ, released on 20 July 2017. Available at: http://www. mpf.mp.br/pgr/documentos/pet-6738/view. 32 Despite severely limited public data, Croese (2017, 80-100) demonstrates that Angolan-state agents owned 'private sector' housing projects like Zango, alongside Odebrecht. Such intensity meant that by the time Lula arrived in power, Brazilian diplomats role was more ceremonial: Angola chose Odebrecht to manage Brazilian relations. The company had direct access to President Dos Santos' office and Odebrecht's CEO held annual meetings with him since 1980. Odebrecht did increase operations after the civil war and during the PT era. Given the company's symbiotic relationship with the Angolan regime, by 2014, it was Angola's largest private-sector employee and the leading benefactor of the Angolan state's public infrastructure investments (Alencastro 2019). Odebrecht received about 10% of Angola's public works' budget between 2003 and 2012 undertaking housing projects, petrochemical stations, hydroelectric dams, road paving and supermarket distribution (Alencastro 2014). This increased the company's leverage and made it a de facto extension of the Ministry of Public Works. Nevertheless, Odebrecht wasn't merely a constructor of infrastructure as it also influenced policymaking. For instance, several signature MPLA projects, such as large-scale housing schemes, drew partially on Odebrecht's Brazilian experience. 33 The company also connected Brazilian political advisors with the Angolan government, even paying for a number of political consultants. For instance João Santana, the organiser of President dos Santos and President Dilma Rousseff campaigns, admitted that Odebrecht paid him for Angola work. Odebrecht also orchestrated vast corruption schemes, as revealed by the Lavo Jato investigation. 34 Because of its complete absence of accountability, Odebrecht considered running Brazilian bribery operations from Angola, although this was abandoned for technical reasons (Alencastro 2019). Thus, after its introduction in 1985, and with the Brazilian state's financial and diplomatic backing, Odebrecht' forged engagements in Angola, generating a symbiotic relationship that made it the key political channel between the two countries.
However, in emphasising the importance of Odebrecht, it is important to avoid presenting Angola as a passive recipient, as a subject in Odebrecht's "empire". Indeed media commentaries (e.g. Fellet 2012) and Odebrecht itself have asserted this neo-colonial narrative: one interviewee stated that in the 1980s and 1990s, Angola saw Odebrecht and Brazilian politicians' arrival "as something of a miracle". 35 We refute this, demonstrating that despite the Angolan government's technical limitations, it directed Odebrecht and Brazil's activities to serve its own political interests. For one, the Angolan government had complete say on who got Brazil's oil-for-infrastructure lines-of-credit, making Odebrecht's monopolistic position dependent on Dos Santos' regime (Alencastro 2019). 36 Moreover, Odebrecht's activity wholly formed part of Angola's statebuilding strategy of expanding control across the country and building a loyal middle class (Soares de Oliveira 2015). In paving thousands of kilometres of roads, erecting schools and hospitals, establishing supermarket chains and making agricultural plantations Odebrecht therefore "[acted as an] advanced platform of the Angolan state" 37 ; "O Estado está aqui" (the state is here), often adorned its projects. The case study of Tanzania, through an analysis of its elite-level politics, further demonstrates the importance of companies and African agency in Brazil's ties with the continent.

Brazilian agency
However, the balance between the Brazilian state and corporate actors was different in Tanzania, with Lula's government having a more important role. This partly stems from the fact that Tanzania, as an Anglophone, East African country, was relatively politically, socially and culturally unfamiliar to Brazil. This absence rendered diplomats the key initiating interlocutors, particularly in development cooperation projects. Brazil opened an embassy in 2005 in Dar-es-Salaam and Tanzania followed suit in 2007. This proved important as the second ambassador appointed to the country, Ambassador Luiz, initiated a step-change in relations, starting a number of cooperation projects around biofuels and HIV-AIDS training and education (Uchôa 2010a). Additionally, Brazil chose Tanzania as the location to spend money won in a WTO settlement, for cotton projects in Africa. Brazilian diplomats' also initiated business opportunities. This fitted the wider principle that Brazil should use its "soft power advantages, especially in infrastructure". 38 Consequently, diplomats, after an invitation from Tanzania's President to bring investment, "identified energy, mining and logistics" 39 as focal sectors and approached large Brazilian companies. This level of government involvement meant that Brazilian Ambassadors were crucial in lobbying senior officials and convening follow-up meetings with civil servants and ministers to plan infrastructure projects. 40 The most prominent example here was the Stiegler's Gorge project, where Brazilian diplomats' are quoted or pictured in press releases between 2012 and 2015. 41 High-level visits were also key events building Brazil-Tanzania ties. Foreign Minister Celso Amorim made the first of these in 2009, bringing a trade delegation. President Lula then included Tanzania in a multi-country 2010 trip. He focused on signing cooperation deals and brokering meetings for government and private-sector projects. 42 This included Odebrecht, who reportedly met Tanzanian President Kikwete, the energy ministry and RUBADA, the agency leading numerous dam developments. 43 This demonstrates how under the PT, the state provided strong support for infrastructure companies, using firms to stretch Brazil-Africa relations across the continent.
However, these corporations were not passive. Petrobras started operations before Brazil's 2005 embassy opening. They independently negotiated off-shore exploration licences and block-ownership. Moreover, showing the company's expansionist ambitions and capacity to deliver on the PT's Brazil Policy, Odebrecht was the first to respond to the Brazilian ambassador's call for interest in November 2009, before Lula's state visit. Interview with Senior Odebrecht Official, 2016. regional project. Following a formal MoU granting Odebrecht prerogative to build the dam in 2011, the company stepped in for RUBADA and paid for an initial EIA (Arms on Environment 2014). They also completed deign and feasibility studies and explored financing options, approaching several European governmental export banks and the World Bank, 44 given that Tanzania's outstanding debt to Brazil precluded BNDES finance. Odebrecht also started negotiations on the expansion of Mtwara airport, 45 with this project envisioned as a steppingstone increasing in-country experience and staff to enable dam construction.
Following an invitation from the Tanzanian government (covered below), another Brazilian construction firm, Queiroz Galvão independently signed an MoU for the Mnyera hydropower project. It involves a large dam and a number of downstream hydropower plants. Queiroz Galvão opened a Dar-es-Salaam office in 2012 and began collecting hydrological data. The firm also autonomously produced feasibility and design studies alongside an impact assessment, later approved by the Tanzanian government. Queiroz Galvão also independently sought financing, meeting with the UK export-finance agency amongst others. 46 Like Odebrecht, the company hoped to launch activities with an airport project, this time at Dodoma. Whilst receiving diplomatic support for the dams, negotiations around both airports were driven by each company. 47 Any understanding of Brazil's engagement in Tanzania, what projects it was involved in and why, therefore requires appreciation of Odebrecht and Queiroz Galvão as individual agents that pushed engagements forward through their own investments. Thus, the Tanzanian case exemplifies how corporation and state acted in lockstep in pushing Africa ties. generation (Dye 2018). Abandoning previous commitments, this new minister gave no financial backing to Odebrecht or Queiroz Galvão, nor did he provide the agreements needed by the companies to raise private capital on the international market. Whilst the President and other ministers continued to make supportive statements about the project, progress stalled by 2014. This demonstrates the degree of fragmentation in Tanzania's ruling party, where decision-making didn't flow from centralised plans made by the President, but was rather divided between ministerial fiefdoms (Jacob 2017). Hopes were further dashed by post-2015 President Magufuli, who has dismantled RUBADA and ignored its deals with Odebrecht and Queiroz Galvão. Rather, he has handed contracts for Stiegler's Gorge Dam to Egyptian companies (Dye 2019). It was therefore Tanzanian officials who directed Brazilian companies and its government. With firms' wide-ranging capabilities and the PT's broad support for Brazilian business and development cooperation, the Tanzanian state proved decisive, directing Brazilian actors to specific projects and determining when, or if, they would reach completion.

Progressive development impacts?: Debunking exceptionalism
Thus, whilst very different contexts, both Angola and Tanzania prove the relevance of analysing African elites and Brazilian corporations as, alongside the state, they determined the spread, intensity and success of Brazil-Africa relations. But what of Brazil's claims to exceptionalism and progressive socio-economic outcomes? Both case studies question this. For instance, failure to complete infrastructure projects in Tanzania was widespread: China's Sinohydro, the African Development Bank, USA's Power Africa and UK firm Aldwych concurrently invested in dam studies and lengthy project negotiations without reaching implementation (Dye 2018, Chapter Four). Concurrently in Angola, whilst many projects were completed, few of Odebrecht's projects met expectations of development. From the Capanda hydroelectric plant to the sugar-cane plantation Biocom and the supermarket chain Nosso Super, the majority of Odebrecht's initiatives experienced significant cost overruns whilst delivering less benefit than planned. Because of this, a number of projects were sold to other foreign investors. Most notably, Odebrecht sold the Catoca diamond mine to Chinese LL International Holding BV in January 2018. 51 This undermines the idea of Brazil having a fully progressive influence or unique understanding of "Africa", its governments and business environment, especially given struggles in Tanzania to secure finance or sustained, cross-governmental buy-in for projects. The agency of political elites in Africa is crucial to explaining this. The case of Tanzania shows that its politicians directed projects whilst determining implementation prospects. 52 Additionally, the Tanzanian government followed-up on Brazilian corporate interest and also involved Queiroz Galvão's. Meanwhile, in Angola, the Brazilian government utilised Odebrecht, who remained the Angolan government's preferred interlocutor for Brazilian relations until 2015. The Angolan government used Odebrecht for its political needs, whether undertaking statebuilding infrastructure, spreading illicit wealth amongst key families or maintaining relations with the Brazilian state. Therefore, it was a change in the Angolan government with the end of President Dos Santos' 38-year rule, that primarily caused Odebrecht's more recent collapse in Angola. Political crisis and corruption scandals in Brazil, alongside a decrease in financial support post-PT, compounded the issue, but Angolan actors were the key players here. This political change in 2015 shocked Odebrecht and meant that a number of construction projects and contracts ended whilst from 2016, the company started to be treated like any other.
Our analytical lens therefore shows that Brazil-Africa relations followed typical patterns in the continent's international ties, with African agency helping to shape a narrowly beneficial concentration of investment in infrastructure, mining and fossil fuels that dwarfed development cooperation spending. Brazil's Tanzanian footprint was primarily corporate, even in development cooperation projects given Cotton Victoria's emphasis on enhancing Brazilian agricultural businesses. The largest successful activity was Petrobras' investment in Tanzania's offshore gas. Dams also followed an extractive model. The Tanzanian elite planned them and stood to gain the most, given that only 32.7% of Tanzanians are grid-connected 53 and electricity tariffs make power expensive for the rural smallholder majority. Moreover, the Stiegler's Gorge and Mnyera projects have significant, negative impacts. Both are on the Rufiji River, which runs through the Selous UNESCO World Heritage Site and ends in a delta protected by the Ramsar-Convention. The Stiegler's Gorge Dam is just above the biodiverse heartland of the Selous' ecologicallyoutstanding area. Importantly, the river's annual flood underpins agriculture, fishing and the biodiversity of both protected sites (Dye and Hartmann 2017;Hoag and Öhman 2008). Both dams will therefore carry serious environmental and social trade-offs that would be almost exclusively borne by rural smallholder farmers in marginalised parts of the country. A similar pattern emerged in Angola. The article has recounted Odebrecht's major levels of corruption which saw millions of dollars diverted to the political-economic elite. Moreover, Odebrecht's operations included diamond extraction that replicated the Chinese and Western enclave operations of oil whereby elites accrue a place's wealth without spreading benefits, whether through jobs or direct and indirect revenue sharing. There have also been allegations of human slavery in Odebrecht's Angolan business (Fellet 2012).
Thus, Brazil exported the inequalities of its domestic development and infrastructure mission more than its progressive, poverty-reducing model. Dam building and highway development on both sides of the Atlantic were underpinned by the same bank, BNDES, and involved the same companies. In projects like the Stiegler's Gorge Dam, Brazil became embroiled in similar socio-environmental controversies to those surrounding Amazonian dams given their destruction of globally-important habitat and/or the resources of local agricultural livelihoods. 54 Odebrecht was also conscious of this mirroring. It hoped for a similar intertwining of private companies with the state in African countries. Complete replication didn't take place however, with Angolan elites appropriating elements of Brazilian policies to the extent that they fit the regime's objectives. For instance, Odebrecht hired policy experts, media strategists and campaign managers to legitimise Angola's contested and authoritarian elections. Rather than the initial rhetoric of brotherhood with Africaof advancing the continent's developmentin practice, Brazil's corporations fitted into the political economy norms of the states they engaged with. Far from fulfilling Lula's heralded bettering of Chinese or Western companies Brazilian firms therefore seem thoroughly average. Activities primarily involved the extractive and construction sectors, which are also the largest areas invested in by Chinese companies (Bräutigam 2011;Power, Mohan, and Tan-Mullins 2012). Indeed they are typically the largest economic sectors in the majority of countries in Africa and subject to the biggest international investment flows (Taylor 2014;Carmody 2011). In contrast to the rhetoric asserting an alternative, more progressive pattern of development diversifying African economies, Brazil has, in concert with most international agents in Africa, reinforced long-established patterns of export dependence.

Conclusion
The cases of Angola and Tanzania therefore demonstrate that Brazil's Africa engagement was not exceptional. Lula's dramatic push invoked strident rhetoric about Brazil's benign intentions, its belief in the country's relevance to Africa and the importance of re-making the global system in solidarity with partners on the continent. The PT promised to export the "Brazilian miracle" of poverty and hunger reduction, the expansion of liberal values and political stability. However, this Brazilian vision quickly met a different African reality. Although development cooperation did increase, Brazil's footprint was primarily corporate and focused on the extractive and infrastructure sectors. Similarly, the majority of Brazil-Africa relations followed elite-brokered projects, a pattern common in South-South and Western-based global corporation engagements. Moreover, Angola and Tanzania demonstrate a number of controversial projects primarily benefiting elites at the cost of the environment and average citizen. Brazil-Africa therefore tended to replicate typical economic patterns on the continent, often reinforcing extraversion where investment and economic wealth were focused on primary resources and their extraction to the international market. Brazil's activities additionally parallel a trend in international development, reviewed above, whereby countries in the global north increasingly use aid to support their domestic firms. A rhetorical striving for exceptionalism was not born out on the ground.
A key reason for this was the PT's corporate turn. Although a strong role for corporations in foreign policy represents something of a return to the Brazilian political norm, and whilst not wanting to overstate the schism of 2005s "Mensalão scandal", the resulting alliance for the PT's political survival brought a handful of large corporations increasingly into policymaking. This had notable implications for the direction and implementation of the PT's Africa policy. Although not initiating the Africa agenda, and despite being more reliant on the state in "newer" places like Tanzania, companies notably Vale, Odebrecht and other infrastructure and agribusiness firms, increasingly influenced the distribution of resources and shifted the rhetoric of Brazil-Africa relations during the PT's rule. Emphasis on finance for infrastructure and Brazilian corporate projects increased. This article's demonstration of these companies' agency therefore unpacks "Brazil" as a monolithic entity in Africa and underlines the importance of treating private firms as foreignpolicy actors.
But alongside this domestic engagement of companies, the article asserts the crucial role of African agency in creating these outcomes. The principle of rejecting the notion of African actors as mere passive recipients is well-established in the Africanist, areastudies literature, but has often been overlooked in Brazil-Africa and elsewhere in studies on the emerging powers in Africa (e.g. Burgis 2015). In contrast, evidence here shows that African political elites ensured that Brazil's push to Africa was not exceptional; they incorporated Brazilian actors into their pre-existing patterns of politics and business. This is also true for corruption which, despite being so loudly decried in the Brazilian media, is very much the norm in Mozambique and Angola: Mozambique, for instance, has suffered far more from the 2012 looting of US Bank loans, which it hid from the IMF. Finally, in spite of its record, the perception of Brazil in Africa appears to be relatively positive. The majority of African leaders, for instance, supported Lula in 2016 after his arrest. Equally, anecdotal generalisations amongst Brazil-Africa scholars indicate that where there is an opinion, Brazil tends to be held in a positive light. However, whether the Brazilian state and non-state actors will be able to utilise this in coming decades remains to be seen.