Special Collection: Mega-Projects, Contentious Action, and Policy Change in Latin America The Infrastructure-Extractives-Resource Governance Complex in the Pan-Amazon: Roll Backs and Contestations

Large-scale access and energy infrastructure projects, together with expanding investments in natural resource extraction, pose significant challenges to biodiversity conservation, forest cover, and the defence of forest peoples’ rights and livelihoods across the wider Amazon region. Following a period in which safeguards and forest dwellers’ territorial rights were strengthened under more permissive political opportunity structures, the current period has been characterized by efforts to weaken these protections and to facilitate large-scale private investment in previously protected lands. We describe these investment-based threats to forests and rights, and the nature of regulatory rollbacks in the region. We then discuss some of the ways in which social movement actors have responded to these pressures and the extent to which they have affected the policies driving these pressures on forests and rights. While in prior decades movements were able to exercise mediated influence on policy, at present the channels open to them are mostly indirect, though opportunities for collaboration between movements organizations and rights-defending government agencies do emerge periodically offering channels for mediated influence.


Introduction
Large-scale infrastructure development projects, together with expanding natural resource extraction investments, pose significant challenges to biodiversity conservation initiatives and the defence of forest peoples' rights and livelihoods across the wider Amazon region. Taken together, the number of current and proposed infrastructure projects, including dams, hydropower plants, waterways, highways and feeder roads, plus the settlements and housing for the workers who build the infrastructure, are poised to further transform the world's largest standing rainforest (Fearnside, 2017;Humphreys Bebbington et al., 2018). Natural resource extraction activities, specifically industrial-scale mining and hydrocarbons development, are dependent upon the construction of access infrastructure in the form of roads, railways and port facilities to transport commodities to distant markets. In the Amazon basin, infrastructure and resource extraction investments interact and reinforce each other to produce large-scale landscape and social impacts that urgently require assessment on a basin-wide scale (Latrubesse, 2017). At the same time, the imperative to construct, extract and export has led to social mobilization and protest as indigenous and traditional forest populations are drawn into socio-environmental conflict in a context in which citizen rights are diminished and existing environmental protections directly challenged by powerful interests. This paper considers how far this conflict has affected policies governing such large scale investment in the region.
For decades, large-scale infrastructure development has been equated with progress and development needed by a region that is characterized as suffering from a prolonged scarcity of physical infrastructure (Larde, 2016). In Brazil, successive governments have pursued state-promoted, large-scale development projects -many of which were funded by the World Bank and Inter-American Development Bank (IDB) -to support industrial mining, timber extraction, agricultural production, electricity production and industrialization (Hall, 1989). Over time, investments in large-scale hydropower, road-building and industrial mining and hydrocarbons gave rise to a highly technical, consolidated national construction sector with expertise in, and a penchant for, mega projects as well as an increasing need for such large-scale projects in order to continue operating as large enterprises. In the past two decades, this politically powerful sector has also been a direct beneficiary of regional integration initiatives, International Financial Institutions (IFI) and government funding flows, along with the introduction of public-private partnerships (PPPs) that consolidated expanded business opportunities for Brazilian firms to build infrastructure both domestically and across South America and beyond.
Like Brazil, Andes-Amazon 1 governments also pursued policies to promote and finance the occupation of their respective Amazonian territories, including the exploitation of natural resources, however these processes were slower and often less successful than those of Brazil. The main exceptions to this pattern include eastern Bolivia (Department of Santa Cruz), where colonization schemes and national policy supported the expansion of large-scale cattle-raising and soy cultivation, and oil and gas development (Eastwood & Pollard, 1985;Hecht, 2005), and eastern/north-eastern Ecuador where transnational oil exploration and exploitation, beginning in the 1960s, initiated significant migration into Ecuador's Orient via roads built to support oil development (Lu, Valdivia, & Silva, 2017). In Peru, the Belaúnde government sought to build its own Trans-Amazon highway in the 1960s, and dreamed of relocating Peru's capital to the Amazon (Dourojeanni, 2017). For all the Andes-Amazon governments, the main constraint on developing the Amazon more completely was the lack of state financing.
In the late 1990s, under the leadership of Brazil and with the support of the Inter-American Development Bank, South American governments came together to overcome financial constraints, creating a comprehensive framework to coordinate and finance region-wide investments in infrastructure. In 2000, the Initiative for the Integration of Regional Infrastructure in South America (Iniciativa para la Integración de la Infraestructura Regional Suramericana, IIRSA) was launched in a bid to support a platform of coordinated investments in mega projects along priority development corridors with the aim of positioning South America as a competitor in global markets. The emergence of new markets, in particular China, as consumers of South American commodities, unleashed a series of investments in mega projects not seen before ( van Dijck, 2013).
Following a period of economic and political setbacks, there are strong synergies re-emerging between both national and IIRSA (now named COSIPLAN) infrastructure investments, 2 the expansion of agribusiness and the extraction of minerals and hydrocarbons in the Amazon. This infrastructure agenda has been -and continues to be -driven above all by Brazilian interests, and extends beyond Brazil's borders into neighbouring Andes-Amazon landscapes where complementary investments are coordinated under the banner of regional integration. However, proponents of integration and large-scale infrastructure and extractive activity are encountering significant resistance to their plans. Across local, national, continental, and transnational scales, coalitions of residents, indigenous communities, conservationists, scientists, non-governmental organizations (NGOs) and government officials have mobilised to oppose mega projects using an array of tactics and arguments with a view to slow, redesign, and in some instances stop projects from moving forward. This paper will consider how such movements are mobilising to change governance arrangements at the national and subnational levels.
The paper is organised as follows. First, we present a simple framework through which to understand this most recent round of infrastructure and extraction. This is followed by a discussion of recent and projected patterns of investment in infrastructure and extractive industry, and the links between these two and agro-industrial expansion. We then include a discussion of government efforts to introduce reforms aimed at reducing protected status of land and eliminating bureaucratic impediments to private investment in resource extraction and infrastructure development. The paper then presents a set of experiences in which counter movements of civic actors, sometimes together with actors within the state and judiciary, have responded to these reforms. Counter movement strategies have included direct protest, litigation, research and data analysis aimed at the public sphere and direct negotiation with investors and policy designers. The final section of the paper discusses how these responses have affected the overall form of the infrastructure-extractives-resource governance complex, and considers the extent to which there have been demonstrable changes that offer the prospect of enhancing environmental quality and the status of rights in the presence of large-scale investments.
Research for this paper was carried out from 2016 to 2018. Our methods included an extensive desk-based review of key investments in infrastructure, mining and natural gas in Amazonia; key informant interviews (in person, electronically, and by phone) with representatives from public agencies, non-governmental organizations (NGOs), the private sector, research centres, and private foundations; GIS based analyses of changes in forest cover and extractive industry; and workshops with civil society and social movement organizations held in Brasilia (July, 2017), Lima (November, 2017) and Oslo (April, 2017). The paper also builds on several of the authors' long engagements with social movement organizations and their networks on the themes of natural resource extraction and large-scale infrastructure development in Amazonia.

Roll back and contestation: a framework for analysis
From the late 1970s and into the 1980s and 1990s, newly organised Amazonian groups -supported by transnational activist networks as well as national activists during a period of relative democratic opening -demanded and won greater accountability and protections from IFIs and the private sector, as well as increased national recognition of their territorial claims, albeit more in some countries than others (Hochstetler & Keck, 2007;Brysk, 2000). In particular, the decade of the 1990s, marked by a series of high profile socio-environmental conflicts and transnational activism, created political opportunity to push states, International Financial Institutions and private companies to take seriously accusations of serious human rights violations, deforestation and the destruction of fragile and biodiverse environments. Examples of high profile conflicts include the Kayapó campaigns against hydropower development in Brazil (Hildyard, 1989), the U'wa fight against oil drilling in Eastern Colombia (Arenas, 2007) and the Chiquitanos' opposition to the Cuiabá pipeline in Eastern Bolivia (Hindery, 2013). IFIs responded to pressure by introducing a series of social and environmental safeguards and added indigenous peoples' specialists to their staff. Private sector groups such as the International Council on Mining and Metals (ICMM) followed somewhat later by developing "best practice" guidelines and codes of conduct and encouraging members to adopt them (International Council on Mining andMetals, 2015, 2006). Indigenous and traditional peoples invoked the International Labour Organization's Convention on Indigenous and Tribal Peoples (ILO 169, 1989), and pursued strategic litigation to force states and companies to obtain the consent of affected peoples. Governments moved, albeit slowly, to comply with the Convention's requirements and conduct Free, Prior and Informed Consent (FPIC) with affected populations, and later adopted the UN Declaration on the Rights of Indigenous Peoples (UNDRIP, 2007). This period from the mid-1990s to 2010, then, can be seen as having been characterized by a window of political opportunity in which indigenous groups, traditional peoples and their allies were able to create an environment somewhat more favourable to their concerns. Grassroots groups and civil society, and their allies in government, made some progress towards a social and environmental policy framework that enhanced protections of human rights and the environment.
Currently, there is a sense that not only are these hard-fought for gains once again under risk, but also that the political opportunity structures that had opened and helped make such gains possible in the first place are also now closing across the region. 3 Across the region, a combination of legislative initiatives and presidential decrees are beginning to open up protected areas to the exploration and extraction of minerals and hydrocarbons, restrict the creation of new indigenous territories, reform environmental licensing procedures, restrict consultation processes and allow for the fast-tracking of projects considered to be of strategic national interest. Beginning in 2012, in a context of some commodity price decline and significant slowdown in new investment in resource extraction, community leaders, NGO staff, and environmental and conservation funders have noted a concerted effort by state and private sector actors to rework, dilute and in some cases eliminate social and environmental safeguards related to largescale development investments. These efforts, pursued by both progressive and more conservative governments alike, also seek to relax restrictions around activities that may be conducted in protected areas -be they indigenous territories or conservation areas (Campodónico, 2018;Humphreys Bebbington et al., 2018).
Based on the argument that their countries must maintain competitiveness, social and environmental regulations are cast as overly bureaucratic, burdensome and obstructive to national development goals and plans. In 2016, the World Bank completed a multi-year review of its environmental and social safeguards. While the World Bank reaffirmed its commitment to protecting the rights of vulnerable populations, the new Environmental and Social Framework (ESF) replaces the current rule-based system with a more flexible approach that places responsibility for compliance on country borrower systems. 4 A coalition of human rights, environmental and indigenous rights groups argued that these rule changes will nonetheless mean that borrower standards will be less stringent (Federici, 2015). We describe this concerted effort to rework the rules as "roll back", a movement to reduce or eliminate regulations with the aim of creating a more secure and favourable environment for private investment.
Our use of the term roll back is normative; recognizing that the protections instituted by states to provide more favourable social policy for marginalized and historically discriminated groups is under assault. At the same time, we recognize that what constitutes roll back for some groups, is forward movement for other groups. One example is the Ruralist Lobby (Bancada Ruralista) in Brazil, described more fully later in this paper, which actively works to rescind burdensome regulations and procedures and replace them with legislation that limits state action (Tollefson, 2016(Tollefson, , 2018Verdum, 2016). In this way, the roll back of social and environmental protections is accompanied by a simultaneous rolling out of policy initiatives that work to the advantage of sectoral interests and establish more supportive conditions for infrastructure, agro-industry and extractive development. Roll back is characterised by a deliberate process of loosening rules and protections, creating legal frameworks and norms that promote a more favourable climate for business and investment, and reducing counter-movement access to policy processes.
In the Andean-Amazonian countries a similar phenomenon occurred, albeit with nuanced differences between the countries of the ostensible "pink tide" and those more reliably neoliberal countries . 5 Not only has this "race to the bottom"  involved regulatory changes, it has also constrained civil society and countermovement access to the policy process. If during periods of more open political opportunity structures, movements were able to affect policies in ways that were (Giugni, 2007; Silva, this issue) occasionally direct (i.e. having their proposals directly taken up by policy makers) and more often mediated (with movements and political actors collaborating in policy formulation and implementation), in this period of rollback, movement involvement is at best "indirect" 6 and more often marginal, leaving movements mostly with protest, denuncias in newspapers, litigation and criticism as their primary instruments.

Patterns of investment in Amazonia and their drivers
Agribusiness, dominated by soy, oil palm and beef production, together with energy development (hydropower) and mineral and hydrocarbons extraction drive the economy of Amazonia. These activities are characterized by their scope of production, export-oriented focus and significant levels of (often coordinated) investment from both private and state sources. In spite of recent political scandals, discussed further below, Brazil's model of development based on coordinated, large-scale investments in access infrastructure, agribusiness and extraction remains intact. Beginning in 2007, the federal government launched the first of its Growth Acceleration Programs (PAC 1, 2007(PAC 1, -2010, US$320 billion) to prioritize social, urban, logistics, and energy and infrastructure development throughout the country in support of economic development. Transport and energy investments figured prominently in PAC 1 with major investments proposed for the Brazilian Amazon. Under PAC 2 (2011-2015, US$960 billion) the focus shifted to energy, with electricity production accounting for some 70 per cent of proposed expenditures (US Trade and Development Agency 2016). PAC 1&2, together with sectoral strategic plans under the Ministry of Mines and Energy and the Ministry of Transportation, and the South American Regional Integration Initiative-IIRSA (now COSIPLAN), identified priority projects and in the process provided the framework for complementary investments in extraction and agribusiness to follow.
Andes-Amazon economies have followed a similar model but one in which mining and hydrocarbons figure more prominently. The integration of these economies into new global markets, especially China, has provided new sources of financial capital with important implications for how projects are funded and implemented. Large-scale infrastructure development is not new to the region. However, what is different from the pre-2000 period, is the sheer number of projects going forward, the vast amounts of money involved and the very favourable investment environment in which mineral and agricultural commodities produced by these countries have enjoyed both strong global demand and high prices 7 (Killeen, 2016;Killeen et al., 2007;Bebbington & Bury, 2013;Little, 2014;Charity et al., 2016).
Today's Amazon Infrastructure builds on what was first articulated under IIRSA, which was promoted by the Brazilian government under Fernando Henrique Cardoso and subsequently adopted and expanded on by the Luiz Ignácio Lula de Silva administration. Embraced broadly by South American Presidents, IIRSA provided a framework to coordinate, fund and implement large-scale, complex infrastructure works on a fast-track basis by relocating decision-making about projects to regional level technical groups. At the same time, IIRSA introduced new financing mechanisms, ostensibly to remove the funding shortfalls and bottlenecks that often derail large-scale infrastructure projects.
From the beginning, the Amazon basin has been a central focus of IIRSA with the justification that new East-West routes were needed if the region was to be competitive in a rapidly globalizing world ( van Dijck, 2013). Access infrastructure, specifically highways and waterways, together with hydroelectric power plants, constitute the priority infrastructure projects in Amazonia. In the case of highways and waterways, they form part of multi-modal transport systems designed to move large quantities of commodities over great distances to ports for export. Again, mineral, grain and beef exports to Asia drive access infrastructure investment. Specifically, South America required modern, efficient and harmonious infrastructure to take advantage of new market opportunities in China and India. A series of priority corridors, hubs and projects were identified -with many of the projects recycled versions of earlier proposals for infrastructure projects ( van Dijck, 2008). This high modernist vision (Scott, 1998) hardly varies between the more radical, twenty-first century socialist countries of Venezuela, Ecuador and Bolivia, and the centre-right governments of Brazil, Colombia and Peru. Indeed, what is remarkable is the continuity of visions and rationales for fast-tracking infrastructure and extraction projects, and limiting dissent, across the South American continent (Bebbington & Humphreys Bebbington, 2011).
Energy security, in particular via hydroelectric power generation, has been a particularly important driver of infrastructure investment across the Amazon basin. Recent research indicates that some 246 dams are currently planned or under construction (in comparison with the 191 currently existing), many of which exist in synergistic relationship with expansion of the mining industry in Amazonia (Lees et al., 2016). There are also important reserves of hydrocarbons in yet to be developed areas. Recent mapping exercises identified some 327 oil or natural gas blocks for bidding or under exploration covering some 1.08 million square kilometres. Mining concessions cover a further 1.6 million square kilometresabout 21 per cent of the total area of the basin (RAISG, 2012).
Changes in sources of financing and the mechanisms used to finance infrastructure development in the Brazilian Amazon respond to the emergence of new global actors, national politics and scandals, and pressures from the private special interest groups to resume infrastructure investments. The primary financial flows in support of infrastructure and extraction in the Amazon Basin have been concentrated in three types: International Financial Institutions (IFIs), primarily the Inter-American Development Bank (IDB) but more recently CAF (the Development Bank of Latin America); Brazilian national institutions; and Chinese financial institutions. The IDB continues to provide significant funding of infrastructure (37 projects for US$9.8 billion) while a revamped CAF is close behind with US$8.8 billion. The World Bank is a distant third with just US$792 million in 4 projects. 8 Unlike, the other two IFIs, CAF is a public-private mechanism currently owned by 19 countries in Latin America and the Caribbean and Europe as well as by a series of private banks. National legislative reforms and the creation of public-private partnerships (PPPs) and new funding mechanisms are also central to the advancement of the Amazon Infrastructure Agenda (Coronado, 2017).
Until recently, Brazilian financial institutions played a key role in financing infrastructure projects throughout Amazonia. Among them two such institutions stand out, the state-owned National Development Bank (BNDES) and the Banco do Brasil, which together provide about 60 per cent of rural loans in the country. Loans are concentrated in the soy producing regions of the Amazon and Cerrado. While the vigour of the national financing has decreased somewhat in the wake of the Car Wash (Lava Jato) corruption scandal (see below) and economic slowdown, Chinese financing is expected to increase over the near term as demand for the region's mineral and agricultural commodities is projected to grow over the next two decades, and new business opportunities unfold for direct investment in infrastructure development. Ray and Gallagher (2015) report that China's policy banks are the largest yearly public creditors, providing US$22.1 billion in loans in 2014 -more than the combined lending of the IDB, CAF and World Bank in the Latin America and Caribbean region. A single company, Vale, Brazil's semi-public mining company and a major producer of iron-ore received US$7.5 billion in loans from two Chinese financial entities. In 2015, Chinese President Xi Jinping met with the Presidents of Latin America to discuss business opportunities and pledged US$250 billion in investments over the next decade (Ray and Gallagher, 2017). Trade between China and Latin America has surged from US$12 billion in 2000 to US$285 billion in 2014. In the wake of the Lavo Jato scandal, Chinese firms are moving in to purchase distressed construction firms in the region providing such firms the possibility of not only financing, but also building major infrastructure works in the region. The prospects of increased trade and investment from China is generating concern among Amazon scholars. Fearnside & Figueiredo (2015) found increased forest loss in the Amazon Basin associated with increased trade with China. Chinese consumption of Brazilian soy, beef, iron-ore and timber, as well as Chinese investment in a rail line linking the soy producing Mato Grosso region to Amazon riverine ports for export to China are driving deforestation.
While corruption is mostly about capturing rents associated with resource extraction and infrastructure development, it has, arguably, also helped further drive investment as it creates incentives to expand public investment quickly. Investigations into the Lava Jato scandal involving the Brazilian mega-companies Odebrecht and Petrobras, among others, have uncovered an elaborate, massive corruption scheme implicating a significant number of corporate executives and high-level officials, including elected Presidents, of current and former governments throughout Latin America. 9 In the early 2000s, Brazil's largest construction firms came together to create a cartel to rig bids and contracts on major projects. Cartel members then bribed Petrobras employees and members of the Board of Directors to ensure silence. The recent conviction of Odebrecht's CEO, Marcelo Odebrecht, set off a chain of subsequent inquiries that has revealed illegal payments and influence peddling across Latin America.
The formal complaint against Odebrecht, filed in the United States District Court, Eastern District of New York, citing the Foreign Corrupt Practices Act of 1977, accused Odebrecht of paying out US$349 million in bribes to political parties, foreign officials and their representatives in Brazil in order to gain advantage in business deals for the company over the period . The complaint claims that Odebrecht netted $1.9 billion in exchange for these payments (USA vs Odebrecht, S.A. 2016). While this is hardly the first corruption scandal involving infrastructure works in Brazil, or Latin America, -recall the scandal involving the construction of dams in the 1980s -this case lays bare the systematic organization of political and economic relationships that underlies corrupt practice linked to infrastructure development (Straub, 2015). In spite of new reforms to reign in corruption and increase transparency of public-private firms, such as Petrobras and Vale, it is likely that such networks will find ways to reconstitute themselves with new actors and new flows of capital to direct to priority investments as part of the Amazon Infrastructure Agenda.
Following the Lava Jato scandal, and as oil prices declined, Petrobras was forced to scale back its operations and institute a divestment plan in order to raise funds. This has created opportunities for the development of oil, natural gas and shale gas reserves by foreign investors, among them the world's largest oil companies. Challenges in the hydrocarbons sector are mirrored in the mining sector, where declining mineral prices and uncertainty around proposed reforms to Brazil's mining code under the Rousseff government, stalled investment. 10 Under the Temer government, officials sought to restore investor confidence and interest by creating buzz around potentially attractive mineral deposits and hydrocarbons reserves, demonstrated by Petrobras' bidding auctions and the proposed opening of a massive national mineral reserve, RENCA, in the northern Amazon (though the Brazilian government later retracted the opening of RENCA after a national and international outcry) (Ministério de Minas e Energia, 2017).
Poor experiences during earlier periods of infrastructure development and resource extraction in Brazil have generated a broad literature documenting forest loss linked to highway construction, mining, and infrastructure (Hall, 1989;Killeen & others, 2007;W. Laurance, 2012;W. F. Laurance, Goosem, & Laurance, 2009;Alamgir et al., 2017;Dourojeanni, Barandiarán, & Dourojeanni, 2009). Research on dam construction and hydroelectric plants points to harmful impacts that extend well beyond the footprint of the project and the cascading effects for forests, rivers and tributaries, and local populations (Fearnside, 2014(Fearnside, , 2006Finer & Jenkins, 2012). Multiple infrastructure investments, i.e. roadbuilding linked to hydroelectric projects, can drive migration of people looking for employment and economic opportunity, land speculation and occupation, and practices of deforestation of land to signal possession.
The packages of investments being proposed and pursued frequently overlap with protected areas, specifically legally recognized indigenous territories and territories under consolidation. The Amazonian Network of Georeferenced Socio-environmental Information (RAISG) has conducted cartographic analysis of these threats, and concluded that across the Amazon Basin, nearly all protected areas, both indigenous and traditional peoples' lands, as well as lands held under conservation, are threatened by some form of hydro-power/waterway development, mining, hydrocarbon, and road-building. In many cases, these protected areas are facing a combination of the above (RAISG, 2012). 11 At the same time, government proclivity to defend protected areas and indigenous territories from encroachment by development initiatives is at best equivocal, with only some agencies within government showing any particular predisposition for such defence (Fearnside, 2017).
Given these antecedents, and these currently overlapping geographies, it is perhaps not surprising that, despite strong public support for large-scale infrastructure across the region, there have been strong, negative reactions from indigenous groups and traditional populations, conservationists, socio-environmental activists and academics. In some cases (see later) these responses have influenced debates around IFI funding standards, social and environmental safeguards, and the need for greater transparency and accountability. In part to respond to the rising challenge, and improve the credibility of integration initiatives, IIRSA took on a name change and introduced new mechanisms to deepen efforts to catalyse investments. In 2010, IIRSA became the South American Planning and Infrastructure Council (Consejo Sudamericano de Planeamiento e Infraestructura, COSIPLAN) and took over responsibility of the portfolio of IIRSA projects.

The rollback of regulatory frameworks
Governments throughout the Amazon basin have responded to declining international prices and weak investment by further deregulating mining and hydrocarbons sectors and increasing output . The move to create conditions that are more favourable for private foreign investors, including nontraditional investors, includes legislative and executive manoeuvres to create new sub-soil markets by opening up previously restricted areas for investment in exploratory activity (Campodónico, 2018). The competitiveness argument is frequently deployed to justify increased investment in energy and transport infrastructure, especially infrastructure related to bulk transport systems of agricultural commodities.
Both the Lula and Rousseff governments vigorously promoted energy development as a means to propel economic growth. With the bulk of projects located in Amazonia, rules and protocols such as consultation with affected indigenous and traditional peoples, and environmental licensing were relaxed to avoid project delays (Fearnside, 2017). At the same time, in order to move the Amazon Infrastructure Agenda forward, the government began slowing down the formal recognition and titling of indigenous lands (Gonçalves et al., 2014). The Rousseff government also issued Provisional Measure No. 558, a unilateral decree, resulting in the downsizing of seven federal Conservation Units in order to advance four dam projects on the Madeira and Tapajós Rivers in the Amazon Basin.
An initial survey of proposed legislation in Brazil's National Congress revealed some 140 proposals aimed directly at affecting land rights and/or indigenous rights over natural resources and cultural patrimony. Two proposals seek to allow mineral exploration in Indigenous Territories (Laws 1610/96 and 37/2011) while other legislation proposes to establish limits on indigenous and traditional peoples' rights under the justification of "the Union's relevant interest". In the case of the latter, the legislation would reform Article 231 of the Federal Constitution.
Behind the raft of proposed legislative and normative changes has been the powerful Ruralist Lobby now organized as the Frente Parlamentar Mista da Agropecuária. The Ruralist Lobby is a major force in Brazil's national politics and its influence in the National Congress had grown from 20 deputies in 1994 to 227 deputies (out of 513) and 27 senators (out of 81) prior to the national elections of October 2018. The rise of the Ruralist Lobby reflects the increasing political power of the agribusiness sector and the networks of economic and political interests and relationships that sustain the Amazon Infrastructure Agenda (Gonzales, 2017). It also reflects how increased Chinese demand for soy and beef may be driving a reformulation of the national political settlement around natural resources in Brazil -from a settlement in which organized urban and rural workers had more presence towards one in which they and their interests are increasingly marginalized while those of the Ruralist Lobby are far more powerful (Fearnside & Figueiredo, 2016; see Bebbington et al., this issue, on political settlements). The Ruralist Lobby has changed legislation regulating the granting of public lands to private farmers and reduced farmer environmental obligations on holdings of up to 1500 hectares in the Amazon. In 2011, the Ruralist Lobby promoted a law weakening the federal environmental protection agency IBAMA, and its power to impose fines on farmers for illegal deforestation. The following year, the Ruralist Lobby successfully pushed through a reformed Forest Code that provided amnesty to farmers who had illegally cleared forests up to the year 2008 and lifted the requirement on farmers to restore deforested lands.
The Ruralist Lobby together with extractive industry have sought to alter the protected status of Indigenous Territories and traditional peoples' lands in order to facilitate infrastructure development and extractive investments. The rules that currently protect Indigenous Peoples' territories are seen as obstacles to the Amazon Infrastructure Agenda and a serious brake on private mining investment. The proposed Constitutional Amendment #215/2000, for instance, seeks to transfer authority for recognizing and demarcating indigenous land claims, and the creation of protected areas, from the Executive Branch to the National Congress (Gonçalves et al., 2014). While the amendment has been stuck in committee stage in the Chamber of Deputies, it is now ready to be included and voted on in the coming agenda of the full Chamber. At the same time, the new President Bolsonaro 12 has made clear his antipathy to titling indigenous people's lands and his belief that they should be made available for resource extraction or agro-enterprise.
Not dissimilar initiatives are evident in the Andean-Amazonian countries. In Bolivia, the MAS/Morales government revised laws to permit hydrocarbon exploration in protected areas arguing that the country needed to develop its reserves to sustain the broad process of social change (Campanini, 2015;Fernández-Llamazares & Rocha, 2015). Environmental analysts argue that the new law is part of a larger strategy to both reduce government oversight of protected areas as well as create new sub-soil markets for hydrocarbon development in previously restricted areas. This has been contentious, as in the case of government efforts to introduce fracking into the Tariquía Park in the Department of Tarija in Southern Bolivia, which has led to social conflict as a coalition of communities, NGOs, and urban-based groups, supported by the Governor, have called on MAS to desist.
In Ecuador and Peru, extractive industry activity is now permitted (albeit under certain conditions) in areas with indigenous peoples living in voluntary isolation (Humphreys . Governments have moved to reduce periods established for consultation and environmental licensing, reduce the amount of information required, and in some instances, suspend regulatory requirements in order to advance specific projects. The most important social and environmental safeguards impacted by the relaxation of rules are environmental and social assessments and consultation processes. In Peru, the relaxation of environmental standards has reduced the role of the Ministry of Environment in creating protected areas and diminished its supervisory role of the mining sector related to environmental compliance (a role that had only been won several years previously). During 2018, the Government of Peru has sought to advance a new hydrocarbons law which critics argue will increase company tenure rights (lengthening contracts to at least 60 years), weaken territorial rights, restrict the Ministry of Environment's authorities, and reduce royalties (Campodónico, 2018). 13 Similar situations have occurred in Ecuador and Bolivia. In Ecuador, the Ministry of Environment now falls under the Coordinating Ministry of Strategic Sectors, which is also responsible for the Ministries of Mining and Hydrocarbons. In Colombia, subnational governments have lost control over the supervision of activities in their regions.
One tactic used by Amazonian governments to fast track infrastructure and resource extraction initiatives has been to invoke national strategic interest, national security and public interest arguments. In Peru, the García government negotiated a bilateral agreement with Brazil to supply about 6,000 MW of hydroelectric power from six Amazonian dams over three decades in a proposal to create a new energy artery (The Economist, 2011). Both leaders argued that the controversial agreement was in the strategic national interests of their countries.
The Brazilian legislature is considering several legislative proposals which propose to use paragraph 6 of article 231 of the Federal Constitution of 1988, concerning "the Union's relevant interest," in order to establish limits on the ethnic and territorial rights of indigenous peoples, as well as other traditional communities in the Amazon and elsewhere (Verdum, 2016). In a growing number of situations, the government has invoked the "security suspension", suspensão de segurança, in order to quell public dissent and to fast track projects. In this way, politicians and government officials use legal levers to neutralize social and environmental protections (Fearnside, 2015).

Protest and counter movements: between mediated and indirect effects
When read in the round, extraction and infrastructure expansion in the Amazon basin, has combined to erode the rights of local populations and adversely affect their natural resource based livelihoods. This expansion impacts property rights as reflected initially in cases where concessions overlap with pre-existing community, territorial and other land rights and then subsequently in displacement driven by unfair land sales or forced eviction. Another set of impacts affect rights to consultation, in those cases where free, prior and informed consent (FPIC) processes are either avoided or short-circuited. In the worst cases, the rights of certain groups to existence are also impacted. This is especially the case of indigenous populations living in voluntary isolation, though it has also been the case for environmental defenders. Indeed, the failure of governments to consult and include local populations in decision-making about development investments is seen as the main proximate driver behind increased levels of socioenvironmental conflict (Global Witness, 2017). Communities, organizations, environmentalists and human rights activists who criticize and protest infrastructure and extractive projects are regularly denounced as agents of foreign governments, anti-development, opportunistic agitators and anti-patriotic (Bebbington & Humphreys Bebbington, 2011), and increasingly protest has been criminalized (Civicus and PWYP, 2016). Governments have taken to threatening civil society organizations with loss of their legal status and funding. Organizations have been audited and subjected to other forms of harassment. 14 Environmental defenders, leaders and community members are intimidated, roughed up, arrested, and murdered. Global Witness reports (2017, 2018) note that more environmental activists were killed in 2015, 2016 and 2017 in Brazil than in any other country in the world for (though on a per capita basis Honduras and Nicaragua are more dangerous still). In 2015, 48 activists -45 of them from the Amazon -were killed according to Brazil's Land Pastoral Commission.
And yet, even in such difficult contexts there is evidence that some protests and mobilizations have led to impact on policies, as well as on projects whose sheer scale elevates their approval to the level of a policy decision. Here we discuss several such cases. As we do so, we bear in mind Giugni's (2007) distinctions between direct, mediated and indirect effects of protest on policy (see Silva, this issue). We suggest that while there are some experiences of mediated effects -i.e. where movements and policy makers and framers work together to design policy related to Amazonian rights and environments -most demonstrated influence on policy has been indirect in nature. Namely, "movements through protest and/or the generation of favourable public opinion place an issue on the policy agenda and political actors subsequently address them without movement involvement in the policy process" (Silva, this issue). We first note an illustration from Bolivia, then develop fuller examples from Peru and Brazil.
In Bolivia, lowland indigenous opposition to the construction of the TIPNIS highway led to violent confrontation between marchers and the police. Subsequent government statements that the highway would be built regardless further emboldened the opposition. At first, denying the need to conduct a consultation process with affected populations, the government later reversed itself and agreed to conduct an FPIC process. However, MAS' awkwardly orchestrated consultation process only furthered perceptions that the government was violating the rights of local indigenous populations while reinforcing the idea that the right to be consulted is constitutive of one's rights as a citizen. The TIPNIS protestors forced a response from the Bolivian government, and has also enrolled a number of international observers and actors. How far the protest directly affected policy remains to be seen: to date, it has led to a de facto policy of still no completed road, though the extent to which this response stands the test of time is unclear given President Evo Morales' statements that the road will go ahead following review.
In Brazil, the conflict over the Tapajós River and the proposed construction of the Tapajós hydroelectric complex induced the emergence of a coalition of indigenous communities, environmentalists, scientists, international allies and government lawyers who argued that the government must follow due process and consult with the indigenous Munduruku. When the government responded that it did not have a protocol for consultation, the Munduruku proposed their own consultation process and implemented it. Contentious action around the proposed investment also led to deeper scrutiny over the environmental licensing process and the preparation of environmental impact assessments, in particular the need for more rigorous assessments that consider the cumulative and synergistic impacts of multiple dams combined with other infrastructure such as waterways and highways. Considered an important piece in the consolidation of a soy transport corridor, the Ruralist Lobby responded to the challenge by requesting a security suspension in order to allow the project to proceed without any further environmental license. In 2016, the project was indefinitely shelved after Brazil's environmental protection agency, IBAMA, withdrew permits for the São Luiz de Tapajós dam.
How resilient this postponement of the project will be is, however, uncertain in the light of similar experiences elsewhere in Brazil. In the 1980s, a network of actors organized under the "Xingu Campaign" were similarly able to stop the building of the Kararaô hydroelectric project (Comissão Pró-Índio de São Paulo, 1988). When in the 1990s the government proposed changes to the design of the project, a further decade of pressure, protest and negotiations continued, with certain government agencies also aligning with the criticisms of the project (Funai, Ibama, ICMBio, and the Ministério Público). However, by 2007, the Lula government included the Belo Monte dam project in its first Growth Acceleration Programme, and by 2010 the project had been contracted to a consortium of Brazilian companies. The implication is that protest was able to delay the project, and to have mediated effects on its design, but was not, ultimately, able to stop the project. Once the political opportunity structure shifted under Lula, the scope to question the project was reduced.
Yet by December 2017, the context appeared to shift again, when Brazil's Executive Secretary of the Ministry of Energy and Mines, Paulo Pedrosa, announced the suspension of all dam construction in the Amazon. In his declaration, Pedrosa acknowledged the increasing complexity around environmental licensing and financing mega-dams in the wake of a series of controversial projects, among them, the Belo Monte Dam. He went on to say that the country would pursue new directions to meet energy needs and would explore the development of clean renewable power (O Globo, 2018). The announcement, however, does not necessarily signal an end to dam building in the Amazon but may instead represent a strategic pause as the government moves to address financing problems and redefine protocols. Indeed, only one week prior to the Ministry's announcement, the National Energy Agency (ANEEL) approved technical viability studies for the proposed Jatoba dam on the Tapajós River, another massive dam project. Any such partial ambivalence about dams is absent in neighbouring Bolivia, where President Evo Morales reaffirmed his intention to move forward on a series of proposed mega dam building projects through 2025 (Medina Candia, 2018;Los Tiempos, 2018). The hydropower complexes of Chepete-El Bala, Rositas and the binational projects of Rio Madeira and Cachuela Esperanza will all export energy to Brazil. 15 This raises the possibility that changes to policy on hydroelectricity in one Amazonian country may simply displace investment to another country, given that the increasing integration of energy networks can allow for cross-border trade in electricity.
A third illustration draws on the history of contestation, reform, roll back and contestation that runs from the Camisea natural gas project from the 1990s/early 2000s through to the consequences of the Bagua protests in 2009. The Camisea project elicited a campaign to either block or modify the design of the investment, given its overlaps with and proximity to primary forest and indigenous territory, including of peoples living in voluntary isolation. Arguably because of the significant involvement of the Inter-American Development Bank in the project and the international nature of the campaign, protest was able to work with project planners and change its design such that four new protected areas were created or strengthened, 16 and greater protections were given to the Territorial Reserve of Nahua Kugapkori. Shortly afterwards a law protecting indigenous people living in voluntary isolation was passed in 2006.
While the Camisea mobilizations secured changes at the scale of national policy, and did so through interactions between actors, government and financiers, the Bagua protests of 2009, while having an impact on policy also, did so in a way that was much more indirect and related to the intensity of the protest and the harshness of the violence that accompanied it. In this case, indigenous people in the Peruvian Amazon were protesting against a series of decrees that together constituted a substantial effort at roll back by the Alan García government in 2008-2009. These decrees were to affect the legal recognition of indigenous communities, the titling of their lands, the forest law, and the promotion of extractive industry investments, inter alia. These efforts induced a massive Amazonian mobilisation in Bagua resulting in the death of 34 people in large part because of the government's repressive response to local protest. In the aftermath of the violence, the state and indigenous people did engage in working groups, which contributed to subsequent laws: the Law of Prior Consultation and the Forestry and Wildlife Law, each in 2011. Though not addressing everything that protesters had wanted, these laws did include some of their concerns, in particular for community forest management; legal recognition of community forests; and the requirement that prior to signing contracts with hydrocarbon companies, the government would have to undertake a process of prior consultation with indigenous peoples.

Concluding comments
Arguably, the suspension of an investment on the scale of the São Luiz de Tapajós dam is of such significance that it constitutes an institutional change produced by protest; certainly the freezing of investment in dams as a sector would be an important policy change (albeit fragile and reversible). However, in other instances, the impacts of protest are limited to the project level while wider "rules of the game" and dominant ideas remain unchanged. Indeed, at the level of ideas, significant parts of the infrastructure and extractive industry sectors continue to view indigenous people and protected areas as an obstacle to a particular investment rather than the basis of a different way of thinking about development. As Walker notes, "The unfortunate fact is that arguments about environmental and social consequences of hydropower carry little weight among the powers that be in most of South America." 17 Notwithstanding significant levels of campaigning, protest, litigation and conflict, the tendency of the last decade has been for governments to push forward aggressively with investment in infrastructure and extractives. The most prominent institutional changes have been oriented towards facilitating investment, not towards subjecting it to more intensive socio-environmental oversight. In this sense, while protest might drive institutional change, declining rates of growth and revenues -and fear of future declines -have been more significant drivers of new institutional arrangements. The reform efforts of the Temer government are the extreme example of this sort of response, but as we have noted, there is evidence of similar tendencies, albeit less aggressive, in Bolivia, Peru and Ecuador -as well as in post-peace Colombia and Venezuela. The election of the Bolsonaro government in Brazil will only deepen this trend, and the Bancada Ruralista is already pushing for further weakening of oversight.
The implication is that the interaction between protest, progressive policy change and roll back is almost Sisyphean in character in the sense that it involves ebb and flow over time, and that to stabilize any policy change requires continued exercise of force in order to prevent roll back. And as in the myth, where the balance in this ebb and flow lies depends entirely on constellations of strength and power. These in turn depend on the relative capacities of different political actors, the organizational capacities of different state, private and civil society agencies pushing for regulation or deregulation, and the political economy and ideational context that frames opportunities for these actors . As commodity prices decline, and investment falls, these contexts become more challenging for movements and those public and civic agencies that seek to defend the environment and community rights, and give ideational resources to groups seeking to promote investments and constrain citizenship rights. In these contexts, the scope for direct or even mediated impact on policy declines, even when some allies within government remain. The current moment in the Amazon is characterized by these contextual conditions, making it that much more challenging for movements to catalyse policy change absent new ideational resources that may once again expand their political opportunities. 9. In Peru, for instance, the scandal forced the resignation of President Pedro Pablo Kuczynski in addition to investigations involving past presidents Alejandro Toledo, Alan García and Ollanta Humala. 10. For example, the price for one key export, iron ore, fell from US$180 per metric ton in 2011 to US$50 per metric ton by the end of 2015. Today the price has somewhat recovered but is still well below its peak. 11. We refer to all indigenous peoples recognized and pending claimed territories across Amazonia as Indigenous Territories (ITs). The term "traditional peoples" refers to a mix of different populations including quilombas, long-time riverside communities and forestbased extractivist communities whose presence dates back to the early twentieth century. 12. Bolsonaro was elected as this paper was being revised, and so is not discussed hereexcept to note that all indications from his statements, and from early manoeuvrings of the Frente Parlamentar da Agropecuária are that the rollback tendencies discussed here will accelerate significantly (see https://agencia.fpagropecuaria.org.br/2018/10/30/fpaaponta-projetos-prioritarios-para-votacao-ainda-neste-ano/, accessed November 1, 2018). 13. The Ministry of Energy and Mines has disputed this interpretation: see for instance, https://gestion.pe/economia/mem-responde-mitos-difundidos-proyecto-nueva-ley-hidrocarburos-245703. 14. In 2016, Acción Ecológica was threatened with closure by the Correa government. In Bolivia, the Fundación Tierra, CEDLA and CEDIB have all been subjected to audits and threatened with closure. Bolivia revoked the status of Danish funder IBIS in 2013. 15. The situation in Peru was different again. Between 2008 and 2011, the Alan García government had given concessions for five pre-investment studies to go ahead for five separate hydroelectricity projects in the Peruvian Amazon (Inambari, Pakitzapango, Tambo 40, Tambo 60 and Mainique I), all to be built by Brazilian companies (Odebrecht, OAS, Furnas, Eletrobras, Camargo Correa) and to produce electricity for Brazil. Yet after a national and international campaign that emphasized free prior and informed consent and national sovereignty, none of the projects ultimately proceeded. 16. These were: the National Sanctuary Megantoni, the National Park Otishi, the Machiguenga y Yanesha Communal Reserve, and the Manu National Park. 17. Robert Walker "To Keep a River Running" http://www.earthisland.org/journal/index.php/eij/article/to_keep_a_river_running/.