Is Brazil a Geoeconomic Node? Geography, Public Policy, and the Failure of Economic Integration in South America

Brazil has been labeled an anchor country, a leading area, and a regional power. Yet, even before the crisis triggered by Operation ‘Car Wash’ began, several scholars had called into question Brazil’s driving role in regional integration, stressing political challenges and economic weaknesses that hindered closer relationships among the South American countries. More optimistic research tends to concentrate on initiatives and visions of Brazil’s regional leadership, with lesser focus on obstacles and implementation. We develop the concept of ‘geoeconomic nodality’ to assess Brazi’s impact on South America and shed light on the structural sources of economic fragmentation, namely geographical conditions and their interaction with public policies. A geoeconomic node is the core of economic networks in a geographically delimited system. The flows of the system ’s units are focused on the node, enabling it to transfer impulses for development – and reflecting what the concepts on anchor countries, leading areas, and regional powers suggest. Our findings show that long distances, physical barriers, the maritime orientation of core zones of population and economic activity, and the poor state of transcontinental infrastructure reduce Brazil’s geoeconomic nodality. Resource nationalism, volatile public policies, and fluctuating exchange rates contribute to this structural mix, so that the prospects to overcome the obstacles imposed by geography appear dim.

supported the internationalization of Brazilian firms throughout the region.
Brazil's political ambitions reach beyond the Southern Cone, as exemplified by its role in the establishment of the Union of South American Nations (UNASUR) in 2008.
Yet, some scholars have voiced skepticism regarding Brazil's influence and impact on South America (BURGES, 2005;DOCTOR, 2013;MALAMUD, 2011MALAMUD, , 2005MALAMUD and GARDINI, 2012;SPEKTOR, 2010). We contribute to this debate by offering a perspective from political geography. This allows us to assess de facto interaction (instead of mere declarations and initiatives). It also sheds light on a usually overlooked factor: the impact of geographical conditions upon the relationships between Brazil and its neighboring countries, especially regarding what we call 'geoeconomic nodality'.
A geoeconomic node is the core of economic networks in a geographically delimited system. The flows of all units that are part of the system are focused on the node, enabling it to transfer momentum for development. Geoeconomic nodality is a sine-qua-non condition for anchor countries, leading areas, and regional powers to exist at all, and for regional integration to progress.
The concept of geoeconomic nodality leads to an assessment of structure, not agency. We are not saying that studies that deal with agency -revealing, for example, how different interest groups shape Brazil's approach toward the region (CASON and POWER, 2009) or how Brazil strategically defects from leading integration (KRAPOHL, 2019) -are without merit. On the contrary, we acknowledge that agency influences some geographical conditions in the long run.
It is important for public policies, which our analysis covers too. However, an analysis of geoeconomic nodality should shed light on aspects neglected by mainstream research, thus complementing the state of the art.
The time-sensitive data of our analysis concentrates on the period from 2008 to 2017, which arguably displays the arc of Brazil's regional influence from zenith to nadir. By chance -or not -it also coincides with the life span of UNASUR, from its foundation to its abandonment by most member states.
This article consists of five sections. First, we summarize the literature on emerging economies and regional development, with particular regard to Brazil, and introduce the defining features of geoeconomic nodality. Second, we examine Is Brazil a Geoeconomic Node? Geography, Public Policy, and the Failure of Economic Integration in South America (2020) 14 (2) e0004 -4/39 regional patterns of trade and investment so as to capture Brazil's nodality. Third, we analyze the impact of location and physical barriers on the relations between Brazil and the rest of South America. Fourth, we assess the state of the region's infrastructure for energy and transport. Fifth, we explain how public policies have influenced Brazil's prospects as a geoeconomic node, and then offer conclusions.

Conceptual framework
The term anchor country encompasses emerging economies that play a central role in the development of regions of (sub)continental scale. The criterion that determines whether a state is an anchor country is its share of the regional economic output. Anchor countries are supposed to be economically more diversified and to possess a share of industrial production above the average of the Global South. As a consequence, they are more competitive than other developing countries and serve as regional growth engines: prosperity in anchor countries is expected to lead to regional prosperity; recession in anchor countries to regional recession (STAMM, 2004). The World Bank (2009) has advanced a similar understanding in its World Development Report. It labels countries that are economically more developed than their surroundings as leading areas -Brazil and South Africa, for example. The key argument is that leading areas offer a high density of economic activity, and developing countries have to reduce distance and division vis -àvis leading areas so as to benefit from this density. Spatial disparities will eventually decrease due to economic impulses generated by leading areas. Remote areas cannot be successful in isolation: their development requires links to the greater national and (sub)continental economy.
Research on regional powers starts with the idea that these states are influential (ØSTERUD, 1992). To qualify as a regional power, a state must be closely linked to its region in cultural, economic, and political ways (NOLTE, 2010). Hence, regional connectivity, which is a mere assumption in the anchor-country concept, becomes a condition for regional powers. Sometimes, this defining feature has been reduced to a criterion "for distinguishing and classifying different types of Sören Scholvin, Andrés Malamud (2020) 14 (2) e0004 -5/39 regional powers" (FLEMES and NOLTE, 2010, p. 07). Yet, if a regional power is not closely linked to its neighbors, why should one expect it to have an impact on them?
We show that regional connectivity is one of Brazil's key shortcomings.
Brazil does not serve as a geoeconomic node of South America because it hardly connects with many of its neighboring countries. Research on the political economy of regional powers stresses the relevance of such connectivity. Kappel (2012) argues that regional powers achieve dominance over other states through unbalanced trade relations, the provision of capital, goods, and services. Regional powers are industrial cores. Sophisticated production is concentrated there, whereas more basic and standardized production is relocated to the regional peripheries. Regional powers are thus tied to their spheres of influence in value chains that allow for regionally dispersed and integrated production.
So how do the three concepts -anchor countries, leading areas, and regional powers -apply to Brazil? Basic trade data, IIRSA, and political commitment suggest that Brazil has played an important role in South America.
The country accounts for about 50 percent of the regional economic output.
Others -mainly Argentina in the automotive sector -are bound to the Brazilian economy in value chains. Their prosperity depends on Brazil's. The mid-2000s were marked by considerable efforts toward institution building within the Common Market of the South (Mercosur), especially to allow for developmental cooperation (RAMANZINI JR. and MARIANO, 2018;RIGGIROZZI and TUSSIE, 2012). Saraiva (2010) argues that economic and political integration in South America has been critical to Brazil for decades, as it guarantees foreign political autonomy visà-vis the United States and creates a fundament for Brazil's global aspirations. At least prior to the end of Dilma Rousseff's presidency, Brazil did indeed pursue a proactive regional agenda, as summarized by Lima (2014).
Against the backdrop of these developments, Milani, Pinheiro and Lima (2017) advance the concept of 'graduation', arguing that the rise of states such as Brazil comes along with a shift from an orientation to the Global North to deeper regional integration that follows a long-term vision 1 . consists of local projects: the bituminization of a road along a few dozen kilometers, the construction of a regional airport or the upgrading of a border post. Eighty-three percent of IIRSA's projects are single-country projects.
Bilateral projects reach 16 percent, and only one percent is multilateral (COSIPLAN, 2017). The Inter-American Development Bank (IADB, 2008) concludes that the dominance of single-country projects has kept IIRSA from fulfilling its mandate to foster regional integration.
Against the backdrop of regional cooperation being more rhetoric than reality, regional integration had come to a dead end even before Brazil's present economic and political crisis began 2 . Burges (2005) refers to the low level of ______________________________________________________________________________________________ 1 Milani, Pinheiro and Lima (2017) also refer to the global scale, where graduated powers become rule makers. We limit our analysis to the regional level. 2 Here and further below, we refer to the sharp decline of Brazil 's GDP growth rate, with the economy contracting in 2015 and 2016. The related political crisis brought about the impeachment of President Rousseff, the interim Temer presidency, and the tense 2018 election, which led to the controversial Bolsonaro presidency.

Sören Scholvin, Andrés Malamud
(2020) 14 (2) e0004 -7/39 intraregional trade and argues that South America lacks a business sector like the European one that could be the driver of deeper integration. Taking up Mattli's idea that regional integration depends on leading powers that not only coordinate policies but also "ease distributional tension by acting as […] 'paymasters'" (MATTLI, 1999, p. 04), Malamud (2011Malamud ( , 2005 highlights that Brazil does not contribute disproportionately to regional integration. Doctor (2013) shows that economic asymmetries and institutional deficits hamper deepening integration in Mercosur. Malamud and Gardini (2012) suggest that regionalism has peaked and delivers diminishing returns. Spektor (2010) argues that already during the proactive Lula era, Brazil preferred regional cooperation to remain fragile and minimalist.
These problems have been reinforced by the more recent economic downturn and political instability. Suffering from severe budget constraints, Brazil could hardly make the necessary financial contribution to regional integration today, regardless of the fact that doing so is not on the agenda of President Jair Bolsonaro, who appears to have little interest in Mercosur -except for its role as an instrument for a trade agreement with the European Union.
All of the above are economic and political challenges to Brazil's role as an anchor country, leading area, and regional power. Debates on Brazil's role in South America would also benefit from recognizing the relevance of geographyespecially because an analysis of geographical constraints highlights the tremendous gap between the rhetoric and reality of regional integration and development (MALAMUD, 2018). Geographical factors interact with public policies (both being conditions or independent variables) to determine the economic impact of Brazil on the region (which is the outcome or dependent variable). In order to uncover this causal relationship, this article focuses on Brazil's prospects as a geoeconomic node.
Among the publications we have reviewed, only Viola and Lima (2017) mention geographical conditions. They point out that physical barriers such as the Andes and poor transport infrastructure obstruct regional value chains in South America, whereas the ease of maritime transport favors them in Southeast Asia. Unfortunately, the two authors do not go into details. As a side note, the gravity model -widely used by economists who study regional integration such as  Bown et al. (2017) and Frankel, Stein and Wei (1995) -is based on distance as an independent variable. According to this model, the intensity of bilateral trade is expected to correspond to the size of the gross domestic product of the countries in consideration and distance between them. However, such research often applies a simplistic understanding of distance, measuring it in Euclidean terms -how many kilometers country A and B are apart. We conceptualize distance in a more complex and thus more realistic way. For example, Colombia is not economically close to Brazil, although they share a common border. Euclidean distance between Brasília and La Paz may be relatively low, but the two cities are further apart than, for instance, Seoul and Singapore are.
Our analytical framework derives from the lifework of the political geographer Saul Cohen. Cohen's (1963) analyses begin with an investigation of location and physical barriers. For instance, he points out that after the Adams-Onís Treaty signed by Spain and the US in 1819, Florida remained a peripheral part of the US for several decades, marked by swamps hindering any kind of development. The peninsula blocked rather than enabled the US expansion into the Caribbean. Cohen (1963) also examines core areas of population and economic activity as well as the connections between them, contextualizing these phenomena with regard to location and physical barriers: Florida turned into a stepping stone for the US into the Caribbean when infrastructure was built there and population density increased.
In this sense, Cohen's research revolves around "the interaction between, on the one hand, geographical settings […] and, on the other, political processes" (2009, p. 12). He sees the "differentiated political space" as a result of the interplay "between the physical and the cultural [meaning man-made] environment" (COHEN, 1957, p. 05). Further to that, Florida's role in the US expansion into the Caribbean exemplifies that political space is "shaped by two forces -the centrifugal and the centripetal" (COHEN, 2009, p. 34), with the former driving separation between territories and the latter promoting territorial unity. Centrifugal and centripetal forces comprise both naturally given and man-made geography. In other words, states shape their own material surroundings to a certain extent -for instance by building infrastructure to overcome physical barriers.
Cohen's position holds geography as a holistic science that covers all factors that matter, including many that political scientists associate with constructivist and institutionalist theories of international relations. Ours, meanwhile, is an approach that narrows the analytical scope down to material conditions that exist in geographic space. Rephrasing the conditions central to Cohen, our analysis focuses on four determinants of geoeconomic nodality: 01.
location, distance, and physical barriers, as fundamental factors that provide a first indication of the extent to which a region can be tied to a geoeconomic node; 02. the distribution of the population and economic activity, which impinges on the cohesion of a region and thus on geoeconomic nodality; 03. infrastructure for energy and transport, which constitutes the key manmade, centripetal force that helps to overcome distance and physical barriers; and 04. public policies, which condition the impact of the aforementioned determinants by boosting or hampering geoeconomic nodality.
Brazil's prospects as a geoeconomic node are also influenced by

Economic ties
A pre-crisis survey of Brazil's largest transnational companies revealed that each of the most internationalized among them -for instance Banco do Brasil, Odebrecht, and Vale -were present in five to eight South American countries.
These enterprises were (and still are) also active in numerous Asian, European However, in addition to the aforementioned shortcomings of IIRSA's projects, not everything financed by the BNDES contributes to regional integration: for example, Odebrecht benefited from such credits provided for the expansion of the subway network of Caracas.
Counter intuitively, the involvement of Brazilian firms abroad has increased in the course of the last years. This does not apply to all Brazilian companies, but there appear to be many that rely on foreign markets -in Europe, operations throughout the region so as to circumvent market access restrictions (or, in less frequent cases, to access resources for consumption in Brazil).
Efficiency and strategic asset seeking, which would result in sophisticated regional value chains, remain marginal.
Brazil increasingly participates in global value chains as an exporter of primary-sector goods, rather than fostering regional value chains that would drive industrialization. According to figures provided by Viola and Lima (2017), manufacturing accounts for not much more than 10 percent of the Brazilian GDP (down from 16 to 18 per cent in the second half of the 2000s) and slightly less than 40 percent of all exports (down from a peak of almost 60 percent at the beginning of this century). The two authors furthermore point out that Brazil is a closed economy -measured by the share of exports in GDP and the share of foreign valueadded in gross exports. The latter stands at 11 percent, as compared to a global average of 24 percent.
Brazil does serve as a considerable supplier to the region, which becomes apparent in trade in automobiles and electronics -more than half of Brazil's corresponding exports go to South America (K AMIYA, 2014) 3 .
Because of its closed economy, Brazil is not, however, a major importer in the regional context. Production inputs are domestic or from overseas ( IAPADRE AND TAJOLI, 2014). Chen and De Lombaerde (2014) also find that Brazil has increased its exports to the region but not its imports therefrom. Regional value chains as a driver of industrialization should follow a reverse pattern, with the geoeconomic node being fed by production inputs -that is, intermediary goods -from the region, further processing and then exporting them globally.
Such a hub -linked to the spokes via backward linkages -would be necessary for South America to copy the Far Eastern model of development Brazil is at least a candidate in this regard, possibly serving as a dual hub together with Argentina. Such a vision is, however, still far from reality.
The regional activities of Petrobras exemplify these limits to Brazil's role as a geoeconomic node. Petrobras is the major buyer of natural gas from Bolivia. This is pure resource-seeking investment. Before Brazil's current crisis, the concentrate on the Brazilian pre-salt resources -a decision whose magnitude was reinforced by the corruption scandal, which practically cut Petrobras off from international credit markets. Petrobras hence sold its assets abroad in order to be able to invest in Brazil 4 .
Further to that, Brazil's role as a geoeconomic node must be spatially delineated in a precise way. The literature on Brazil as an anchor country, leading area and regional powers refers to South America, as do Brazil's official foreign policy documents, so we follow suit. As Figure 01 and Bown et al. (2017), contradicts this interpretation because trade among the South American countries is slightly more than the model predicts. What matters to this article however is that our benchmark -the concepts of anchor countries, leading areas, and regional powers -leads to different expectations. Revisiting these concepts against the backdrop of the gravity model is an interesting endeavor for a different article.

Sören Scholvin, Andrés Malamud
(2020) 14 (2) e0004 -13/39     On the Chilean side, the slope has a far higher grade so that the road consists of a long series of switchbacks to make the descent. Sometimes the path must be closed in winter, when snow blocks its ends and the threat of rockfall is considerable.

Transcontinental infrastructure
At the beginning of this century, a report by the IADB (2000) showed that trade in South America remained low because of insufficient transport The term highway may be misleading for readers used to transport infrastructure in the Global North. In South America, so called highways outside major cities are paved (or unpaved) two-lane roads, whose condition ranges from sufficient to miserable. Transport by rail is even more difficult because railroad gauges vary. The reason for this is that railroads were initially built for bringing primary-sector goods to the coast. They were not meant to serve as a means of transnational or even transcontinental transport. In the La Plata region, tracks were built with different gauges from the mid-nineteenth century onward so that they could not be used by invading armies. Even the few existing railroad corridors do not allow for efficient transport. The one from Buenos Aires to São Paulo is 300 kilometers What is more, Palestini and Agostinis (2015) observe that Brazil's initial commitment to the rehabilitation of transport infrastructure in South America reflected the efforts of the Fernando Henrique Cardoso presidency to physically integrate the entire Brazilian territory in the context of the Avança Brasil agenda.
Putting corresponding projects in a continental instead of national frame eased access to financial support from organizations such as ECLAC and the IADB. In other words, the national focus of IIRSA perfectly matches Brazilian interests, which have little to do with regional integration.
One may suggest that trade in services constitutes an alternative to trade in goods for South America -not only because trade in services is not hampered by insufficient rail and road networks, but also because more and more services have become tradable, offering considerable opportunities for emerging economies to provide them to their respective regions, as Turok and Visagie (2019) explain with regard to South Africa. Viola and Lima (2017)  With regard to trade in electricity, the potential of further integration looks equally dim. Experts of the Latin American Energy Organization (OLADE, 2003) argued at the beginning of this century that Argentina and Brazil possess a realistic transfer potential of 5,000 megawatts, which equals only 3.3 percent of Brazil's installed capacity. Argentina and Chile were expected to reach transfers of not more than 500 megawatts; so were Brazil and Uruguay. Colombia and Ecuador as well as Ecuador and Peru, were predicted to transfer 250 megawatts bilaterally 10 . In more recent publications, OLADE (2013OLADE ( , 2012 develops scenarios of regional cooperation and highlights according benefits, but these documents fall short of indicating steps already taken or likely to be taken so as to realize these wishful aspirations. As Scholvin and Betz ______________________________________________________________________________________________ 10 These calculations exclude the output of bi-and multinational hydropower stations.

Public policies
At Zibechi (2012) argues that South America -especially Bolivia, Paraguay, and Uruguay -were transformed into economies subordinated to Brazil, into stepping stones for the globalization of these sectoral champions.
Another key means of the BNDES to boost the regional standing of  What is more, some regional states pursue policies that are counterproductive to regional cooperation. approach is also innovative insofar as it draws attention to neglected causal factors: geographical conditions that interact with public policies. We contend that such causal factors are an important piece of the puzzle that is Brazil 's role in South America.
The distinctiveness of our approach becomes particularly clear when compared to Milani, Pinheiro and Lima's (2017) insightful article.
They find that cohesion among domestic elites has often been insufficient for Brazil to become a regional leader or paymaster, a graduated power in their terminology.
In contrast, our analysis demonstrated that there are structural constraints to Brazil's graduation or regional hegemony -if one considers connectivity and influence in the economic sphere as vital elements of hegemony. We argue, therefore, that Brazil's 'graduation dilemma' also rests on structure -that is, geoeconomic nodality -and not merely on agency.
Brazil's interaction with its region is much less expressive than what one expects from an anchor country, leading area, or regional power.
Certainly, the internationalization of Brazilian firms appears to be concentrated on South America. Brazilian construction companies have been involved in important infrastructure projects, benefiting from financial support granted by the BNDES.
Still, Brazilian investment in the region is market and resource-seeking. Except for a few cases such as Argentina's automotive sector, it hardly generates value chains, which could trigger development in the neighboring countries. While Brazil is an important trading partner of the Southern Cone (except for Chile), trade with the Andean and Caribbean countries is very limited. Brazil's role as a source of FDI remains negligible, with the exception of Paraguay and Uruguay. Regional exports account for a very low share of the regional GDP. Therefore, we would not even claim that Brazil is the geoeconomic node of the Southern Cone, albeit its relevance there is higher than in South America as a whole.
The concept of geoeconomic nodality rests on four factors (location and physical barriers, the distribution of the population and economic activity, infrastructure for energy and transport, and public policies from many of its neighbors. The Brazilian highlands and the landscape configuration of the coast implies that the Brazilian economy is marked by a maritime orientation. Second, the Brazilian population concentrates close to the Atlantic Ocean, compounding the impact of large distances to the neighboring countries. Third, although transport corridors could help to overcome physical barriers, South America suffers from a lack of railroads and roads that bind the regional countries together. IIRSA serves as a means to advance projects of national scope rather than fostering physical integration across borders. Energyrelated infrastructure does not allow for close cross-border ties either. Fourth, public policies interact with natural and man-made geography. This includes the support schemes by the BNDES for Brazil's regional expansion -introduced during the Cardoso presidency, expanded in the PT era and now largely discontinued. Resource nationalism, double taxation, and fluctuating exchange rates work against Brazilian-South American interaction. We admit that assessments of regional integration in South America -or, more generally, in the Global South -must avoid Euro-centric biases. Because of the different trajectory of economic relations, and the smaller and less developed markets in the Global South, such a comparison would miss the point. Even prosperous emerging economies, such as Brazil until 2014, lack the financial means that initially allowed Germany and France to drive European integration as paymasters (MATTLI, 1999). At this stage, regional integration beyond Europe and North America should focus on developing physical conditions -in particular infrastructure for energy and transport -that improve connectivity and foster regional interdependence. Our analysis demonstrated that there are major obstacles in this regard in South America. Yet, considering budgets constraints due to recurrent economic crises and chronic political instability, as well as the re-orientation of Brazil's economic and foreign policies under Bolsonaro, it is unlikely that regional economic integration will receive any meaningful boost in the coming years.