The Ruins of a Steel Mill: Planetary Urbanization in the Brazilian Amazon

Abstract

This article studies the socio-environmental history of a steel mill in Manaus, Brazil, the largest city in Amazonia. The Companhia Siderúrgica da Amazônia Sociedade Anônima, SIDERAMA, an integrated steel mill created in 1961 and liquidated in 1997 after a convoluted history of mismanagement, was part of private and public efforts to colonize the Brazilian Amazon through industrial urbanization. Exploring its impacts at a diversity of scales, the article presents it as a historical case study of planetary urbanization. The steel mill polluted the city of Manaus, and resource extraction to feed it with materials and energy disrupted environments and societies deep into the Amazon rainforest. This process underscores the global reach of urbanization during the second half of the twentieth century and the need to rethink the role of Amazonia during the Anthropocene.

The encounter between the Amazon River and the Rio Negro is one of the main touristic attractions of Manaus, the capital of the Brazilian state of Amazonas and the largest city of the Amazon rainforest with more than two million inhabitants. The dark and clear waters of two of the world’s greatest rivers meet but run in parallel, creating a striking visual effect. As they enjoy this unique sight, the dozens of tourists that experience “the meeting of the waters” by boat every day could not fail to notice another unusual feature of the landscape: nearby, on the banks of the Rio Negro, a massive, old industrial chimney presides over the area. Behind it, still visible, are more ruins of a recent industrial past. Forest growth has overtaken parts of it, while informal urbanization claims other sections. This is what remains of the Companhia Siderúrgica da Amazônia Sociedade Anônima, SIDERAMA, an integrated steel mill created in 1961 and liquidated in 1997. The rise and fall of SIDERAMA and its relationship with the urbanization of the Brazilian Amazon are the topics of this essay.

Ghost towns, abandoned buildings, ruins, rubble, and other manifestations of decay have a special place in the history of Amazonia, a region with a long history of socio-ecological fragility and volatile economic cycles. Early accounts by sixteenth-century Spanish conquistadores described large towns deserted by the time of their arrival because indigenous peoples had died from diseases or migrated to new regions to escape the newcomers. ¹ Throughout the colonial period, the mobility of native peoples baffled colonizers, who recurrently encountered empty villages. Transhumance was part of communal responses to colonial conditions. Europeans interpreted it, however, as native peoples’ inability to adapt and settle. ² After the Amazon rubber boom of the second half of the nineteenth century went bust, observers in cities like Manaus were often shocked at the speed with which the city’s so-called Belle Époque buildings were abandoned and crumbled. ³ The ruins of Fôrdlandia, the Ford Motor Company’s agro-industrial venture in Pará, provide the mid-twentieth century striking images of an Amazonian postindustrial rust landscape reclaimed by the jungle. ⁴ Brazilian historian Victor Leonardi’s evocative social history of Airão reveals, moreover, a town born in the late seventeenth-century Rio Negro that was abandoned after three centuries of ups and downs when its last inhabitant left in the 1980s. These ruins often contain complex histories, which can be explored at different scales. ⁵

Because it, too, ended in ruins, the SIDERAMA steel mill provides a unique entry to the entangled histories of national-developmentalism and planetary urbanization in the Brazilian Amazon. ⁶ The steel mill was one of the most ambitious projects ever undertaken in the world’s largest tropical rainforest. Created the year following the inauguration of Brasilia, it was part of a nation-wide attempt to “interiorize development” and of local efforts to finally recover from the long bust that followed the rubber boom in the 1910s. Expectations were that it would lead to a long-awaited transformation of the Amazonian extractivist economy into an industrial one. Manaus, instead of an enclave for import–export activities, would become a productive center and a pole of development. Because it would use supplies from Amazonia and because steel was a base industry, SIDERAMA would create linkages with the Amazonian economy while employing the cheap workforce available in Manaus.

Manaus and Amazonia’s heavy industry and steel, while part of a national-developmental agenda, were in turn tied to global processes. During the twentieth century, urban industrialization, in particular steel, became crucial to nation-states’ establishment of sovereignty. The expansion of control and production was viewed, on one hand, as a way to effectively protect national territories. On the other hand, it asserted a nation-state’s place in an increasingly globalized and competitive international economic order. A steel mill in the urbanizing core of Amazonia, a poor, vast frontier with a long history of competing territorial claims where state power was tenuous, could be the paragon of this project in Brazil. The Manaus industrial pole succeeded in galvanizing explosive economic and demographic growth. Expanding its built environment and economic activity would support the kind of dynamic and permanent settlement that Brazilian elites had dreamed of for decades.

At the same time, SIDERAMA also exemplifies the shortcomings and darker sides of this project. It was a catastrophic economic failure and, above all, it left extreme socio-environmental footprints both within and beyond the city. Attention to the materiality of urban industrialization, especially framed as extended urbanization, illuminates the destructive drive and the unequal effects of this developmental model. Industrialization radicalized problematic aspects of Amazonian urban growth and created new ones. The imposition of built environments in the mold of Western cities implied interventions that pushed popular sectors into informal, ecologically precarious, increasingly stigmatized areas. Rapid urban industrial growth, and the ubiquitous conflicts over urban space it generated, fostered even worse conditions for the urban poor.

Industrialization inverted the logic of frontier urbanization, whereby the city was an epiphenomenon of rural extractive economies and an enclave that facilitated the flow of rainforest commodities. As an industrial center, Manaus itself became the engine of large-scale socio-environmental change. ⁷ During the new urban industrial period, Manaus expanded its tentacles deep into the rainforest and even beyond in search of resources and labor. The quest for the main ingredients of steel production damaged environmental dynamics in Amazonian regions near and far. More dramatically, the city’s appetite for energy had devastating human effects. It crated labor conditions analogous to slavery for immigrants from Southern Brazil who produced charcoal, and potentially genocidal violence against the Waimiri-Atroari indigenous peoples. Industrial Manaus left indelible socio-environmental footprints in the rainforest.

The history of SIDERAMA thus illuminates a key moment in the global history of urbanization. It has all the ingredients for a microhistory of planetary urbanization during the “great acceleration” of human impact on the planet during the second half of the twentieth century. As environmental historians have argued, the growth of cities, especially in the Global South, was “one of the signal characteristics of the Anthropocene.” Fueled largely by immigration from the countryside, cities in the developing world grew at an accelerated rate, with an absolute increase of nearly two billion people between 1950 and 2000, to reach 52 percent of the total population of the Earth. These cities often had ecological impacts at a variety of scales: their poorer inhabitants often live in precarious environmental conditions, and the agglomerations siphon energy and materials and affect ecosystems both near and far. ⁸

The reach of extended urbanization in the second half of the twentieth century to Amazonia, one of the world’s great natural frontiers, underscores the phenomenon’s planetary scale. For the past decade, urban scholars and theorists have been advocating a conception of urbanization that transcends traditional urban centers, framing it as a broad range of spatial transformations in the context of uneven capitalist development. These include the commodification of territory, the organization of space around the operational logistics of capitalist ventures, and the modification of ecosystems at different scales, including the so-called “wilderness,” often with enormous and unequal impact on human communities. ⁹

The perspective of planetary urbanization has inspired work attentive to both empirical detail and conceptual discussion, which emphasizes the processual nature of urbanization. Yet, despite the insistence of leading scholars within this paradigm to define planetary urbanization as a historical phenomenon, there is a dearth of detailed historical scholarship framed around this conceptual apparatus. ¹⁰ Potential explanations for this gap may include historians’ customary mistrust of recent theoretical insights; attachment to periodization, such as an “urban age” and to the distinction between cities as units and other spaces; and uncertainty about types of evidence and methods for multi-scalar analysis. ¹¹

These misgivings aside, the planetary reach of urbanization during the second half of the twentieth century reveals a “critical juncture of globalization,” when spatial relations were rearranged simultaneously in response to global processes, and with long-term consequences and implications beyond usual regional or national scales. ¹² Scholars have long argued that cities present excellent opportunities to explore the links between global and local scales. ¹³ More recently, practitioners of the emerging field of “global microhistory” have shown that the combination of the micro and macro perspectives can provide tools and models to trace the relationships between specific local contexts and global processes. ¹⁴ The history of one steel mill in the urbanization of the Brazilian Amazon signals a culminant point of the process of planetary urbanization and the metastasizing of its worst features. It also provides an alternative account of the place of Amazonia in global history.

Manaus from Amazonian Boomtown to Pole of Development

Manaus was built in territories once inhabited by the Manao peoples, whose name it took. ¹⁵ In 1688, as part of the frontier policy of the Portuguese Empire in the Amazon, colonial officials created the fort of São José da Barra. ¹⁶ The military outpost survived for more than a century, until it became the capital of the Captaincy of São José do Rio Negro in 1792 and then in 1833, after Brazil’s independence, of Alto Amazonas. ¹⁷ Like in much of the world, the advent of steam navigation signaled a historical turning point in Amazonia. ¹⁸ Because of its location in the riverine system, Manaus was selected by the Brazilian government as a hub within Western Amazonia. Heavy national investment incentivized commercial activity and created a permanent fluvial port in the city. Upper Amazonian commodities now reached global markets through the Atlantic and the region received foreign products, technology, peoples, and capital. ¹⁹

In the 1860s, when demand for Amazonian wild rubber exploded in the industrialized world and rubber extraction moved deep into the interior, Manaus was uniquely positioned to benefit. By the 1880s, it had become a booming center within a global trade network connecting Amazonian tree-tappers to docks in New York, Liverpool, and Le Havre, and to industries and consumers everywhere. The city concentrated many of the import–export and outfitting houses that dominated the intricate, debt-based, and exploitative rubber industry, as well as ancillary services and commercial businesses, officials, and workers. ²⁰

Much has been said about the peak of the rubber boom prosperity: contemporary observers, historians, and popular accounts of Manaus circa 1910 have long emphasized the city’s state-of-the-art urban facilities and infrastructure, which rivaled those of any Brazilian city. They also described a diverse population, rich and cosmopolitan cultural life, and opulent buildings. According to one influential historian, slums and poverty did exist, but “the Amazon capital had proportionately less poverty and more fine houses than other cities.” ²¹ Scholars of rubber boom-era Manaus, most prominently those from Amazonia, have since complicated these narratives. Using the archives and methods of social history, they have revealed the precarious lives of most inhabitants, which contrast with the image of the city local elites crafted to tout their civilizing efforts and attract more immigrants and capital. For much of the urban population, the so-called Belle Époque was a mirage. Beyond small parts of the city center, where solid buildings lined famous paved avenues, Amazonian urbanization was a precarious affair. Members of the working class were more likely to live on muddy streets in wooden shacks with thatched roofs. Widespread poverty was compounded by the rainforest’s environmental conditions which, among other issues, facilitated the spread of diseases like malaria. Inequality was extreme. ²²

The collapse of the Amazon rubber economy after Asian rubber plantations took over the market in the 1910s hit Manaus extremely hard. ²³ The bust of this rural industry led, however, to decades-long rural–urban migrations of impoverished workers, which fed Manaus’s demographic growth. From a small town of 50,000 inhabitants in 1900, it counted 75,000 residents in 1920, and 175,000 in 1960. ²⁴ This growth took place almost exclusively in poor neighborhoods, around waterways, and on the outskirts or between the cracks of the formal city. The exception was a brief period during the mid-1940s when the Brazilian government in an agreement with the United States recruited rubber tappers mostly from the Northeast to ensure wartime supplies. In exchange, the United States supported important infrastructural works and social programs in Manaus. ²⁵ Otherwise, new populations challenged the urban order and undermined the modernizing hopes of local and national elites. ²⁶

The modernization of the Brazilian interior, ultimately represented by Amazonia, had become an important aspect of the dominant national-developmental agenda since the governments of Getúlio Vargas (1930-1945). Vargas launched a “March to the West” policy that promoted legislation and created developmental agencies aimed at the country’s interior. The two crucial tenets of future developmental efforts in the Amazon are already visible in these early policies: geopolitics and industrial urbanization. Brazilian ruling classes long considered the Amazon a “void” and in latent danger of foreign and internal threats against national security. During the rubber boom, efforts to impose human technologies and western environments on nature accelerated. This development process caused a major rupture in long-standing regional relationships between local human societies and the natural world. ²⁷ The ideal culmination of this attempted modernization, in Amazonia like in Brazil as a whole, was urban industrialization as an economic, material, social, political, and even civilizing compact. ²⁸

Getúlio Vargas’s iconic “Speech of the Amazon River,” delivered in Manaus in 1940, illustrates the spirit of this agenda. In this speech he argued that Brazil would “conquer and dominate the valleys of the great Equatorial torrents, transforming their extraordinary fertility into disciplined energy.” The history of the Amazon, “from a chapter of the history of the land . . . would become a chapter of the history of civilization.” In the centuries-old struggle “to conquer the land, dominate the water, and subject the rainforest,” moreover, “the city of Manaus was not a minor victory.” ²⁹

Still, postwar Manaus grew as precariously as ever. New poor suburbs with humble houses appeared and expanded. Despite explicit prohibitions and timid efforts to create affordable housing, informal urbanization exploded all over the city. ³⁰ By the 1960s, Manaus showed the signs of half a century of deterioration. Its most recognizable landmark was no longer one of the lavish constructions from the rubber era. It was instead the Cidade Flutuante, a floating favela (slum) built on stilts and rafts on the Rio Negro, strategically located in front of the city’s market and close to the city’s poorly maintained port. ³¹ Nearby, the famous Opera House, Teatro Amazonas, languished, chastised by the rainforest weather. ³² The city’s university, one of the oldest in the country, closed in 1926 but for some isolated units. ³³ The city’s decline and pathos became dominant aspects of regional identity.

Local politics in Manaus were controlled by populist leaders, who often resorted to anti-democratic strategies and factional political violence, and by a commercial elite that remained relevant because of the region’s consumer dependence on trade. ³⁴ Through its influential guild, this merchant elite constantly advocated for measures to reactivate the Amazonian economy. It insistently sought governmental incentives and subsidies. The content of these demands reflected their urgency to launch a new developmental model based on industries that processed Amazonian raw materials rather than on commodity exports vulnerable to boom and bust cycles. This would have the additional benefit of incentivizing the construction of enduring infrastructure, which in turn would encourage the increase of the urban population under better conditions. In 1960, Amazonense entrepreneurs who, following national and global trends, reimagined their role from merchants to industrialists, formed a new guild: The Federation of Industries of the State of Amazonas. ³⁵

The main request of the local elite was the implementation of a Free Trade Zone for imports and exports, which had been proposed in the late 1940s and approved by the Brazilian Congress in 1957. Although, initially, it would mostly benefit the established merchant houses, the project’s ultimate goal was to industrialize Amazonia. ³⁶ A first step in the creation of an industrial pole, subsidized imports were part of a larger geopolitical strategy. The establishment of a Free Trade Zone to rival those in Iquitos (Peru) and Leticia (Colombia) was meant to help long-suffering, isolated, and politically abandoned Western Amazonia protect and occupy territories along its extensive international frontier. ³⁷

These specific demands coincided with a crescendo in the broader drive to develop Amazonia. Part of that drive was the construction of the new modernist capital of Brasília under Juscelino Kubitscheck (1956-1961). Built in the middle of the cerrado (subtropical savanna), at the center of an ambitious projected road network, Brasília would serve as a launching pad to conquer the country’s interior, of which Amazonia represented the largest part. ³⁸ When a military dictatorship overtook the Brazilian government on March 31, 1964, it radicalized efforts to modernize the Amazon, powered by the convergence of geopolitical paranoia and developmental hubris. ³⁹ The regime of 1964-1985 pursued a myriad of initiatives with support from local, national, and transnational elites. Amazonia was often framed in explicitly colonial terms and treated as a “spatial fix” for capitalist expansion and national social issues. In the quest for the “interiorization of development,” as Antoine Acker has argued, “the conquest of the Amazon imposed itself progressively as the ultimate objective of this national march toward development.” ⁴⁰

“Operation Amazonia,” launched in 1966, revamped regional development agencies and created new ones, granted subsidies and fiscal incentives to businesses investing in the region, and often directly invested public funds. ⁴¹ The most publicized—though largely failed—program of the dictatorship is the Trans Amazonian highway and the agricultural colonization scheme created on its margins. Much more successful, however, were efforts to promote the growth of Manaus, whose modernization, planners believed, would encourage a more intensive and permanent settlement of Amazonia. The creation of the Manaus Free Trade Zone in 1967 and the Industrial Pole in 1972 fostered impressive demographic and economic growth. ⁴²

An Amazonian Steel Mill

Since the 1930s, Brazilian elites deemed heavy industry, and steel in particular, indispensable for the construction of a modern Brazil. Steel was ubiquitous: it was the spine of modern built environments; essential for other industries; critical for infrastructure, railroads, and military armament; and held great economic promise through the generation of backward and forward linkages. Steel was integral to the developmental agenda of different political regimes, independent of their other ideological leanings. In the national-developmental consensus, a steel industry controlled by Brazilians would improve the country’s balance of trade, help it overcome dependency, shape a modern domestic working class, and fulfill its republican mandate of progress. It was indispensable to Brazil’s self-sufficiency and sovereignty and crucial to Brazil’s capability to determine its own destiny. The nation’s first integrated steel mill, the National Steel Company in Volta Redonda, in the state of Rio de Janeiro, became a powerful symbol of industrial modernity and of a new social and cultural order. ⁴³

If national elites treated industrialization as a key for national sovereignty, regional elites treated it as crucial to the placement of Amazonas, Brazil’s largest state, in the nation. Amazonense elites thus promoted a steel industry in the Amazon as a remedy against their strategic region’s long decline. ⁴⁴ They were inspired by the example of Volta Redonda, built on a former coffee plantation to rescue the Paraíba Valley’s “dead cities” from decay. Enter SIDERAMA, founded in 1961 by lawyer and businessman Sócrates Bomfim. ⁴⁵ The potential of steel, the epitome of heavy industry and urban modernity, was clear to men like Bomfim. It could elevate Manaus’ status among Brazilian cities, create jobs, foster linkages with the rest of the local and regional economy, and support the construction and maintenance of a stable built environment in the rainforest. Its promotors called SIDERAMA a fundamental step in their quest “to fixate man to the land.” ⁴⁶

These lofty objectives were not easy to implement. SIDERAMA’s first years were focused on an aggressive fundraising campaign. Though heavily supported by the federal government—which never owned less than 70 percent of its shares—SIDERAMA pursued private investors through a public relations operation. ⁴⁷ A public stock offerings campaign appealed to regionalist sentiment, with images of an impressive steel mill—not yet existent—and claims that the project would lead the region to “indexes of progress compatible with those in the rest of the country.” The campaign pushed the slogan that “Collaborating with SIDERAMA is the duty of every Amazonian” and promised that “in two years we will produce 30 tons of Amazonian iron for the development of Brazil.” ⁴⁸ By 1975, SIDERAMA counted more than 70,000 individual shareholders from all over Brazil, in part thanks to an effective nation-wide marketing campaign in some of the nation’s most prominent outlets. ⁴⁹ Even the Manaus Free Trade Zone, known in Southern Brazil mostly as a shopping district for discounted foreign goods, became part of the campaign. An advertisement for SIDERAMA in São Paulo’s magazine Veja in 1970 read “Take Advantage of the Manaus Free Trade Zone: Buy a Steel Mill.” ⁵⁰

Political support for the project spanned the political gamut at the local level too, from the populist governor of Amazonas, Governor Plínio Ramos Coelho of the Brazilian Labour Party (PTB) (1955-1959 and 1963-1964), to the dictatorship’s appointed governor and political enemy of Coelho, the conservative Arthur Cézar Ferreira Reis (1964-1967). Different governmental entities also supported and funded SIDERAMA before and after the 1964 coup. Socrates Bomfim regularly traveled to São Paulo, Rio de Janeiro, and Brasilia, the country’s political and financial centers, to secure investments from national, regional, and multinational institutions and household names in global heavy industry. ⁵¹

SIDERAMA secured the real estate for its plant in 1962. The massive, 2.5 million-square-meter plot was located just on the outskirts of Manaus, in a strategic location that would in the following decade become part of the new Industrial Pole. The developing edge of Manaus guaranteed a pool of labor and the resources and services of the city: power, water, communications, and amenities for white-collar workers. The land was deemed capable of withstanding the heavy machinery that steel production would require and the margin of the river stable enough to house a port capable of dealing with large-scale transactions. The immediate vicinity of two large rivers provided direct access to the resources of the rainforest, as well as a relatively direct outlet to markets. ⁵² During the following decade, Amazonian authorities took every high-profile official visitor to Manaus, from foreign dignitaries to Brazilian presidents, to visit the rarely fully functional SIDERAMA in a show of the importance attributed to the steelworks. ⁵³

The construction of the plant began only in 1967, and it took several more years for its inauguration. SIDERAMA symbolically started operations on March 31, 1972, during the visit of president Emílio Garrastazu Médici (1969-1974), the most ironfisted of the Brazilian generals. In an effort to gain his favor, Bomfim announced that the smoke of the chimney would be seen for the first time on that day, the eighth anniversary of the military “revolution.” ⁵⁴ Although massive chimneys stacking smoke over Amazonia were celebrated as emblems of progress, it took six more months, until October 24, 1972, for the plant to produce its first lot of pig iron. ⁵⁵ The inauguration ceremony itself turned out to be a small affair. In his eagerness to please, Bonfim forgot that March 31, 1972, coincided with Good Friday, a day of circumspection in a Catholic country. ⁵⁶ The subdued event foreshadowed SIDERAMA’s lackluster performance: it would never manage to produce a reliable amount of good quality pig iron. ⁵⁷

Two interrelated factors were at the source of many of SIDERAMA’s difficulties: it was planned as an integrated steel mill, and it attempted to exclusively use raw materials from Amazonia. Integrated steelworks create steel from scratch, using iron ores, limestone, and coke coal in blast furnaces, rather than producing it using scrap steel bought from other steel mills. An integrated steel plant would grant the enterprise control over the whole productive process and generate more value without relying on imported pieces. This was key because, for all its ambition, SIDERAMA was a relatively small steel mill, so it could not count on an economy of scale. More importantly, using raw materials from Amazonia would decisively contribute toward the developmental objective of creating regional linkages, fostering integration, and generally introducing Amazonian primary products into the industrial chain. ⁵⁸ On paper and, as it turned out, in reality, importing materials from other Brazilian states or from other nations would also have greatly increased transportation and production costs and would have prevented SIDERAMA from generating revenue from the production of pig iron. In addition to these issues, the diversity of actors and stakeholders involved generated some red tape and a degree of chaos, all of which led to a dysfunctional administration. ⁵⁹

By May 1974, the steel mill was in a deep crisis. Because of a lack of liquidity and unpaid debts, a bank mortgaged some of the company’s equipment and up to 350 workers lost their jobs. ⁶⁰ In retrospect, it never fully recovered. A government audit in 1975 resulted in the Federal Government gaining complete control over the company in the first of many attempts to save its investment, effectively ending Socrates Bomfim’s dream. ⁶¹ Despite explicit support from dictators Medici, Ernesto Geisel (1974-1978), and João Figueiredo (1979-1985), SIDERAMA continued to falter, closing intermittently amid constant demands for additional governmental support. Desperate times brought desperate measures: for all their talk of modernization, authorities brought a priest to bless the steel mill in 1979. ⁶² Not even the supernatural could help: after countless relaunches, attempts to privatize it, stoppages, and re-openings, the Brazilian government finally decided to liquidate SIDERAMA in 1997. ⁶³

Beyond the City

Originally inspired by Henry Lefebvre’s notion of “complete urbanization,” the thesis of planetary urbanization provides a tool to interpret the impacts of SIDERAMA at the local level but also its expanding reach beyond the city proper. Two key elements of planetary urbanization are the interplay of different scales, including the influence of urbanization in faraway places and the key role of environmental change as a function of city growth. If urban political ecology has provided valuable insights into the metabolic flows that constitute life within cities, planetary urbanization can help understand the reach of urbanization into faraway, indeed global realms, through infrastructure and extraction. ⁶⁴

The SIDERAMA plant provides an example of the ecological impact of industrial urbanization at different spatial scales. The plant had been for decades an environmental nuisance in the city itself. Despite the promises of the first years, riverside erosion and riverbank collapses soon threatened to release pollutant chemicals into a critical point of the Amazonian hydraulic system. ⁶⁵ Large government investments to revamp the steel mill in 1979, moreover, did not secure its long-term financial viability but led to a short-lived renaissance during which released toxic fumes contaminated neighboring lands and rivers and affected large portions of Manaus. ⁶⁶ The problem was brought to national attention when Navy officials in neighboring facilities, including a major base, complained that the “insufferable pollution” would “put the life of men at risk.” ⁶⁷

Other industries besides SIDERAMA were involved in pollution scandals in the last third of the twentieth century, the cumulative effects of environmental degradation having drawn more attention to its impact. ⁶⁸ By the early 1990s, Manaus seemed to have reached a tipping point. The quantity, variety, and quality of fish species had been drastically reduced, directly affecting the diet of the city’s poorest populations. Fish still found in the streams and rivers close to the city contained high levels of heavy metals such as copper, zinc, cadmium, chrome, and lead. Some species presented abnormal incidences of tumors. ⁶⁹ Research conducted between 1996 and 1998 showed that the streams of Manaus closest to the industrial district were “in danger of disappearing as a natural ecosystem.” In the most industrialized parts of the city, deforestation was leading to climate change: higher air temperatures and lower precipitation rates. The topography of the streams was drastically modified by erosion and increased sedimentation. The water was not fit for human consumption, and its fauna had been decimated. ⁷⁰

SIDERAMA also had socio-ecological footprints at much larger spatial scales because of the resolve to make it an integrated steel mill that used raw materials from Amazonia. Iron, limestone, and coal, the key ingredients in the steel-making process, were available in Amazonia but had to be transported from far away, using convoluted logistics that generated conflicts between a broad range of local and national actors. In some cases, their poor quality jeopardized the sustainability of the steel mill. And still, insistence in their use extended the plant’s imprint well beyond the air and waters of Manaus.

The steelworks got its iron ore, the main compound for steel production, from a deposit called Mina do Oriente, close to the Jatapú river in the municipality of Urucará, more than 500 kilometers northeast of Manaus. The mine, owned by SIDERAMA founder and first president Socrates Bomfim, was often presented as one of the advantages of the project. ⁷¹ Bomfim’s conviction that the real economic potential of the rainforest lay in its mineral wealth was largely at the root of the whole mad venture. ⁷² In reality, iron ore had to be transported by trucks from the mine to a nearby port, and taken in massive barges to SIDERAMA’s floating deck in Manaus. Transportation was thus costly and complicated. Bomfim nevertheless tried to spin the issue into an asset. As part of SIDERAMA’s “work for Brazil,” he claimed, the plant was “creating towns” and the mining enclaves of Jatapu (iron) and Araras (coal) would “soon become cities.” ⁷³

Even worse, Mina do Oriente’s iron was too rich in goethite, aluminum, and phosphorus. This composition considerably complicated the refining process, making it almost unpractical to produce pig iron. ⁷⁴ When SIDERAMA’s economic failure became patent and it was rescued by the federal government in 1972, state-run audits found that the mineral deposits had never been properly prospected. ⁷⁵ Bomfim then became the target of a broad set of accusations, notably that he had orchestrated the choice of the mine for his own financial gain, causing SIDERAMA to hemorrhage money. ⁷⁶ Still, because the project aligned with the military regime’s long-term interests, and because the “Amazonian miracle” turned the region into a uniquely promising market for steel, SIDERAMA continued to attract shareholders, government subsidies, and state support until neoliberal reforms in the 1990s did away with most state-owned or mixed enterprises. ⁷⁷

Local extraction of limestone, a calcareous sedimentary rock central to steel production, proved to be another example of the negative socio-environmental impact of SIDERAMA’s dysfunction. Limestone is crucial to the process of steel refining and SIDERAMA sought to obtain it from a mineral deposit called Serra da Mulata, 672 kilometers away from Manaus down the Amazon River, in the municipality of Monte Alegre, in Pará. An audit of the site found that SIDERAMA had not done its due diligence there either when it first tested the deposit, but the ore was nevertheless exceptional. ⁷⁸ Exploitation of the site for the benefit of SIDERAMA turned out to be irregular at best and was nearly completely abandoned because of production failures at the steelworks. ⁷⁹ Because it was controlled by the Federal Government, eventually the limestone resources were allocated to another state-owned company, the Commission for the Airports of the Amazon Region (COMARA). ⁸⁰ COMARA’s management of the Monte Alegre mine was disastrous. Field inspectors in 1975 found that they “operated without any concern for safety,” using explosives “without any technical guidance,” and wasting as much as 30 percent of the mineral to mismanagement. The inspector reported that Monte Alegre had become a garimpo, as informal mines are called in Brazil, the mine’s rising output never making it to the books and likely sold under the table to Army officials. ⁸¹ Government attempts to reactivate SIDERAMA later spurred conflicts with COMARA. ⁸² SIDERAMA’s inefficiency again exposed intra-governmental conflict over resources and the predatory extractive agenda of private interests operating in the name of the Brazilian state.

Coke, converted from vegetable coal and used to feed furnaces, was another important raw material for an integrated steel mill. SIDERAMA’s source of coke was Araras, an area of 8,450 hectares only 30 kilometers from Manaus, on the Rio Negro. Though workers found an indigenous cemetery in Araras, plans to use the site and its abundant reservoir of trees moved forward. ⁸³ The employment of hundreds of workers in Araras, who came mainly from the interior of the States of Minas Gerais and Espirito Santo in Southern Brazil, would additionally help to fulfill the industry’s ultimate developmental goal of “occupying, activating, and humanizing” the Amazonian interior. ⁸⁴

Araras turned out to be a horror story. On May 26, 1972, the press in Manaus reported that 450 people were “living isolated from the world, in a regime of slavery” to make coal for the steel industry. ⁸⁵ The main perpetrator was the contractor Gilberto Andrade, who had previously worked for the Belgo-Mineira steel company, which pioneered the use of vegetable coal for steelmaking in Brazil. ⁸⁶ The workers accused Andrade of having brought them by river, and then in trucks, under false promises. More than a hundred workers, with their families, had become indebted to Andrade and his three brothers through their company store and were regularly intimidated by Andrade’s hired thugs (capangas). The local police detail refused to intervene. But when a larger police unit arrived after a tip, they confirmed that citizens from the South were denied access to their earnings, descent accommodations, or sufficient food and medical attention, and they could only leave the site on Sundays and with authorization. ⁸⁷ It is unclear if those responsible for this labor and humanitarian nightmare were ever punished.

Araras’s coal camp also had a harmful ecological impact. The 48,000 tons of coal it produced per year caused extraordinary deforestation. SIDERAMA’s access to subsidized land in 1970 was contingent on an agreement with the government to reforest 1,300 hectares by 1972. ⁸⁸ However, in 1972, a federal inspection reported that the reforestation effort yielded “completely insignificant” results. Out of the 1300 hectares it committed to plant, SIDERAMA had reforested only 15. ⁸⁹ By the 1980s, SIDERAMA had in fact literally carbonized a large part of what it had then named its “forestal project.” New legislation in 1983 imposed strict reforestation demands for ventures like SIDERAMA, which threatened to stop coal production until the company complied. ⁹⁰ Facing another institutional crisis and eager to secure capital injections by the federal government, the steel mill repeatedly petitioned to obtain extensions. ⁹¹ In the meantime, it made arrangements with third parties to create plantations of rosewood and other species of trees. ⁹² But like in other aspects of its operation, SIDERAMA’s coal and reforestation initiative was plagued by corruption and embezzlement. ⁹³ In 1988, the Amazonas office of the Brazilian Institute for Florestal Developmental rejected SIDERAMA’s petitions and demanded that it reforest 20,000 hectares in Araras. ⁹⁴ Its continued failure to do so gravely affected its ability to access governmental support, ultimately contributing to its demise a decade later.

Another form of energy that contributed to SIDERAMA’s crisis was electricity, which the company consumed in great amounts. Its regular inability to pay its electrical bills was a crucial factor in several of its crises. ⁹⁵ Like many other industries based in Manaus, SIDERAMA lobbied to increase the city’s access to electric power through the construction of dams in the Amazon rainforest. The demand for steel that such projects would generate was an added benefit as far as SIDERAMA was concerned. ⁹⁶ The most dramatic consequence of Manaus’ expansion was related to its demand for energy and the impact it had on the regional ecosystem, and above all on the indigenous peoples who lived there. The Balbina hydroelectric dam, built between 1982 and 1988 in the Amazonian Municipality of Presidente Figueiredo, 200 kilometers from Manaus, came about when the region’s power grid was at its nadir. Local elites and state technocracy had demanded the construction of Balbina to help address the industrial pole and the growing metropolis’s insufficient power supply and frequent blackouts. In the 1970s, enthusiasm for the hydroelectric potential of the Amazon rainforest acquired great momentum and the Brazilian government planned dozens of dams for the region. ⁹⁷

By the end of the decade, when SIDERAMA experienced one of its many short-lived resurgences, the steel mill and the dam were perceived as the most important developmental projects in the region. During his visit to Amazonas in 1979, President Figueiredo toured both sites. ⁹⁸ Plans for Balbina were interrupted later that year, however, not for the first or last time. Local politicians and boosters of both SIDERAMA and the dam denounced the delay, complaining it would prevent steel production and sales that were crucial to the mill’s survival. ⁹⁹ When the construction of the dam was resumed in the mid-1980s, SIDERAMA was indeed on its last legs. However, winning the tenders to provide steel for Balbina reanimated the seemingly endless regional optimism for the steelworks. ¹⁰⁰

The choice of the Balbina waterfall in the Uatumã river for the new dam that would feed Manaus power followed plans and studies conducted in 1970 and 1975. It was clear from early on that the choice was problematic given Balbina’s low elevation, the river’s weak streamflow during parts of the year, and potential conflicts in an area that included farmlands and indigenous territories. There were rumors that the selection of Balbina was meant to facilitate mining ventures in the area while justifying questionable compensations to private landowners. The selection was also likely part of political calculations to win the favor of Amazonian elites who demanded mega projects amid an increasingly contested congressional electoral field. In any case, there was little room for oversight of the military regime’s decision. And fierce support from those who benefited from the project—industrial ventures like SIDERAMA, construction companies, financers, and the Manaus constituency—contributed to the “aura of inevitability” that allowed the dam to be built. ¹⁰¹

Tragically, the history of Balbina intersected with the longer history of the relationship between the Brazilian government and the Waimiri and Atroari peoples of the Amazon, whose traditional territory counted close to eight million hectares. ¹⁰² It was a history fraught with conflict, much like the Brazilian state’s broader relationship with indigenous nations. ¹⁰³ The military dictatorship of 1964-1985 intensified this trend, and indeed the final report of Brazil’s National Truth Commission concluded that indigenous peoples were the hardest-hit groups by the regime’s violence. ¹⁰⁴ In the case of the Waimiri-Atroari, researchers have repeatedly concluded that their treatment can only be described as genocidal, resulting in the progressive disappearance of most of their population. ¹⁰⁵ The Waimiri-Atroari, now known as the Kinja, were close to 6,000 people in 1900; by the end of the 1960s, their population had been halved to 3,000; and by the end of the 1980s, this number had been decimated to close to 300. Manaus and its predecessor, the Forte da Barra, had played an important part in the long history of violence between the Waimiri-Atroari and the Brazilian state. ¹⁰⁶

The initial episode related to the dam took place in 1981. A presidential decree reducing the territories of the indigenous reserve to facilitate its construction was repudiated by the Waimiri-Atroari, who were joined in protest by environmentalist groups and sectors of the indigenista progressive Catholic Church. Confidential intelligence reports from the following years show that the Brazilian government was aware of the environmental and social impact of Balbina before flooding the area. Topographical calculations revealed that the affected area would be 3,560 square kilometers instead of the planned 1,600 and that without the construction of a series of dikes an even larger area would be flooded. The Waimiri-Atroari indigenous reserve, already reduced to 49,000 hectares, would in fact only measure 24,800. ¹⁰⁷

The hydroelectric plant took much longer to build than was expected because of excessive costs and corrupt business practices. ¹⁰⁸ But the delays were also related to doubts about the viability of Balbina. ¹⁰⁹ In 1981, for example, another secret report made it clear that authorities knew that energy production in Balbina would not be remotely enough to feed Manaus. ¹¹⁰ Local businessmen reacted badly to any opposition to build a hydroelectric dam. They even wrote a letter to the Brazilian president, signed by industry leaders, that argued that a dam would be “an effective contribution for the occupation of our empty spaces, and to utilize great contingents of workers and construction materials, which will greatly benefit the poor economy of the state of Amazonas.” ¹¹¹ That same year, when President Figueiredo visited Manaus, the state’s governor, José Lindoso, declared in a public speech that the dam was crucial for Amazonian development. A local intelligence agent argued that:

The plant will pioneer the hydraulic potential of Western Amazonia, and has a great economic and psychosocial meaning for the region, taking into account the current surge of development in Manaus, the increase in its population. . . ¹¹²

The enthusiasm for Balbina was in line with state propaganda about the “Brazilian Miracle,” or what Carlos Fico has called “The Reinvention of Optimism,” which often had the modernization of Amazonia at the center of its discourse. ¹¹³ In 1981, for example, one of the justifications for continued work on Balbina despite the considerably slow progress was that the sale of the excess energy it produced would provide significant income in foreign currency. ¹¹⁴ However, just a year before Balbina was inaugurated, a pamphlet published by the regional electric company made it clear that the dam was expected to fulfill less than half of the city’s demand for electric power. ¹¹⁵ Combined with authoritarian politics, capitalist urbanization imposed its logic on the hinterland even if it bordered on the irrational from any other perspective.

It was also known that existing vegetation around the dam would cause pollution once the area was flooded. ¹¹⁶ In 1988, when the gates opened, water flooded between 2,600 and 4,000 kilometers more than was previously estimated. The resulting shallow lake created a patchy landscape that dried during parts of the year, becoming a breeding ground for disease-carrying mosquitoes. Decomposing vegetation contributed to elevated water acidity levels that killed thousands of fish. The lake also liberated hydrogen sulfide gas, which created a foul smell, and even methane gas, making Balbina a contributor to climate change. ¹¹⁷ Thus, the demands of urban growth effectively transformed a region deep in the Amazon rainforest into a factor of planetary environmental change.

These severe environmental consequences paled in comparison with the damage inflicted on the Waimiri-Atroari and the region’s farming communities. The dam fragmented the indigenous lands and forced up to one-third of the remaining Waimiri-Atroari population to move. Decisions about their relocation were not coordinated with the indigenous elders, but with younger members of the community, likely under coercion or cooptation, and after years of systematic violence. ¹¹⁸ The attack against the Waimiri-Atroari took place in the name of industrial urbanization. The quest for national development and sovereignty and the transformation of the regional economy via urban modernization justified environmental and human destruction in the Amazon rainforest. The longer history of Manaus shows that disruption to Amazonian ecosystems and people was always at the core of the project to build modern environments in the region as if the two could not coexist. Industrial Manaus grew by destroying everything in its way.

Epilogue: Ruins

The explosive growth of Manaus since the creation of the Industrial District in 1972 radically transformed the city and its role in the rainforest. The city quickly became a juggernaut. Its population more than tripled between 1960 and 1980 to 635,000 inhabitants, and it reached one million ten years later. ¹¹⁹ In 1960, more than 63 percent of the population of Brazilian Amazonia still lived in rural settlements. ¹²⁰ By 1991, however, 58 percent of the region’s inhabitants lived in cities, and Manaus, with a population of 1.4 million in 2000, surpassed for the first time Belém do Pará, the old colonial capital in the Amazon estuary, becoming the region’s metropolitan center. ¹²¹

Manaus’s economic rise to regional primacy came even earlier and was also sharper. The implementation of Manaus’s Zona Franca (Free Trade Zone) in 1967 and Industrial Pole in 1972 replaced the traditional rural extractive activities at the core of the region’s economy with urban industrial and commercial activities. The gross domestic product (GDP) of Manaus grew at an average yearly rate of 13.1 percent during the 1970s, surpassing an already impressive national average of 8.2 percent. In the 1980s, often called Brazil’s “lost decade,” the national economy slowed down to an average 1.8 percent GDP yearly growth rate, but Manaus’s growth rate was 5.6 percent. ¹²² This growth exacerbated socio-spatial conflicts and environmental degradation.

Paradoxically, this growth took place while the city’s most ambitious industrial project, SIDERAMA, failed. The success of the Manaus Industrial Pole relied on the import and assembly of manufactured pieces rather than heavy industries. Assembly electronics industries gained the most traction and became Zona Franca’s leading sector in the following decades. They made the best of what the Manaus free trade zone had to offer. Their products’ value per weight justified the transportation costs, they enjoyed fiscal exemptions and subsidies granted to industrial products, and they could use Manaus’s existing export/import networks and available unskilled—and cheap—workforce. ¹²³ In the late 1960s, companies from traditional manufacturing centers in São Paulo and other cities in Southern Brazil began moving to Manaus to assemble televisions, radios, video cassette players, fax machines, watches, and other technological marvels of the time. ¹²⁴ By the end of the 1980s, electronics directly accounted for more than 37,000, close to 50 percent, of all jobs in the industrial pole. Besides companies from other parts of Brazil, Manaus managed to attract major multinational factories in the sector, including those of brands with international recognition such as Philips, Sharp, and Honda. These products were then distributed to other parts of Brazil, and, exceptionally, also exported to other countries. ¹²⁵

Beyond the direct creation of jobs, Manaus’s Zona Franca and industrial pole stimulated different aspects of the urban economy, especially construction, services, and government jobs. However, the political-economic regime remained very much that of an enclave with very few linkages with the regional economy. The rapid growth of Manaus, moreover, generated countless conflicts, from increased exploitation of workers and labor mobilization to several disputes over space and informal urbanization. These two trends eventually converged and informal urban associations and industrial worker unions became the regional vanguard of the opposition to the dictatorship in the mid-1980s. Informal urbanization and worker mobilization, moreover, remained a major source of conflict throughout the rest of the twentieth century in Manaus. ¹²⁶

The failure of SIDERAMA is therefore illustrative. By the mid-1980s, despite periodic rushes of enthusiasm, it was clear for most discerning observers that the steel mill would never be a functioning enterprise. In the middle years of the decade, as the political situation of the dictatorship became more fragile, metal workers mobilized against the regime, further threatening the company’s bottom line. ¹²⁷ And yet, SIDERAMA continued to operate, however, precariously and unexpectedly. As the war against drugs became a ubiquitous facet of urban life, Manaus’s police authorities and steel mill administrators devised a new collaboration: the furnaces of SIDERAMA would be used to incinerate confiscated marijuana and cocaine. The smoke emitted by the chimneys of the steel mill, once hailed as a symbol of a new industrial era, was now a reminder of a major aspect the urban crisis of the postindustrial age. ¹²⁸

In 1997, the Brazilian government decided to give up on SIDERAMA and the steelworks finally closed for good. What followed was a series of invasions during the 2000s. First, sporadic trespassers looted the remaining facilities to sell them on the black market. Gradually dozens of families, mostly immigrants from the Amazonian interior, turned them into a living space. By the second decade of the twenty-first century, SIDERAMA’s “City of Steel” lay in ruins, the paradigmatic modern built environment reclaimed by the rainforest. From its ruins emerged a favela. ¹²⁹

In his classic guide to the discipline of history, many decades ago, Marc Bloch warned that exploring failures is indispensable to understanding historical processes: “the unsuccessful act is one of the essential data of human evolution.” ¹³⁰ In another classic essay, the pioneering environmental historian William Cronon showed that close attention to what lies behind ruination has the potential to illuminate otherwise ignored stories about social and material entanglements. ¹³¹ More recently, in her celebrated “ethnography of global connection” about the place of Indonesia in capitalist globalization, the anthropologist Anna Tsing argued that global processes of capitalist expansion often produce “rubble and mud, the residues of success and failure”; it is through these residues that we can begin to think about these histories. ¹³²

Not by coincidence, scholars of the Amazon and other forested borderlands of South America have emphasized the analytical potential of destruction and ruination. In junctures of extreme socio-environmental change such as deforestation, distinctions between production and destruction often collapse in the practice and subjectivity of local peoples, who frame them according to myriad economic and cultural considerations. ¹³³ Similarly, ruins, rubbles, and debris, “ruptured, fraught objects that denaturalize the present,” can in fact be deeply meaningful and even productive spaces for local peoples. ¹³⁴ SIDERAMA was a failed project, part of an ambitious scheme of industrial urbanization. Planners expected the steelworks and urbanization to affect change in the Amazon rainforest, but urban growth acquired a dynamic of its own, creating ruins and destruction at a diversity of scales. Now an abandoned steel factory at the edge of a massive city in the middle of the world’s greatest rainforest, SIDERAMA is reclaimed both by the forest and by the informal city. What remains of it, in all the landscapes and waterscapes it transformed and destroyed, in the written record, and in the stories we tell, can also denaturalize the past and help us rethink the place of Amazonia in planetary urbanization.