Special Collection: Mega-Projects, Contentious Action, and Policy Change in Latin America Mega-Projects, Contentious Politics, and Institutional and Policy Change: Chile, 1994-2017

We think of social movements as forces for change, which frequently passes through policy and institutional reform. Yet just how they affect change all is frequently not addressed. This paper tackles the question in the context of the Chilean environmental movement’s effects on environmental governance and energy policy reforms. Building on work in political sociology it argues that Chilean environmental movement organizations had direct, indirect, and mediated effects on different phases of the policy process. It argues further that a political economy approach to environmental policymaking explains the politics by which they had such effects.


Introduction
Social movements are presumed to be agents of change yet just how they affect change is not easy to determine. To advance our thinking on the problem I focus on the role of the Chilean environmental movement and anti-mega project protest in gradual institutional and policy change in Chile. For institutional change, I analyse a shift from a decentralized inter-ministerial agency in the early 1990s to a Ministry of Environment in 2010 and subsequent related reforms. For policy change I focus on the energy sector, specifically the Law 20/25 of 2013 that mandates 25 percent of energy be produced by non-traditional renewable sources by the year 2020. In both instances I analyse how social movement engagement with the policy process combined with anti-mega project protest affected debates and outcomes. I embed that analysis in a political economy framework that makes explicit the political and economic conditions that inform social movement action and ties to the policy process. To be clear from the outset this is not an argument for the primacy of contentious action or political actors in explaining the reforms. It is about, when, by what mechanisms, and under what conditions movements had some effect at various stages of the policy process.

Tracing the institutional and policy impacts of contentious politics
Building on Giugni (2007) social movements and protest can have direct, indirect, and mediated effects on policy and institutional change. Direct effects occur when politicians adopt movement organization policy proposals virtually unchanged. Indirect effects occur when 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. Mediated effects are those in which after an issue has been placed on the policy agenda movements and political actors collaborate in policy formulation and implementation. Establishing those effects entails tracing the linkages of movement organizations and activists to policy networks involving political parties, legislators, and state managers. It also involves determining the role of protest and contentious action more broadly in getting politicians to recognize a movement's claims and bringing them into the policy process. The focus on policy process means that movements may have effects on only one, some, or all of its various stages: agenda setting, initiation, formulation, implementation, and feedback.
I argue that gradual institutional and policy change in the Chilean environmental issue area involved a shift from primarily direct social movement impacts in agenda setting and indirect impacts in policy formulation to the emergence of mediated impacts. The paper shows, first, that the principal direct effect of environmental movement organizations and protest was placing issues on the policy agenda. The Chilean environmental movement has had a consistent set of demands/policy propositions since Chile's re-democratization in 1990. They often became the environmental agenda of Chilean centre-left presidential candidates.
Second, the paper shows that when it came to policy formulation and implementation movements generally only had an indirect effect on environmental governance regimes and non-traditional renewable energy policy. After social movement organizations had placed those issues on the policy agenda governments acted on them without their input. This was especially true of the first three Concertación administrations (1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005), although mounting protest and public opinion against mega-projects put pressure to keep institutional and policy reform on the agenda.
Third, institutional and policy reforms closer to movement policy proposals occurred in Michelle Bachelet's first government (2006Bachelet's first government ( -2010 and more followed in Sebastián Piñera's centre-right government (2010)(2011)(2012)(2013)(2014). Protests against environmental disasters and socio-environmental conflicts played a major role that shift. They turned public opinion in favour of government action on environmental issues. Politicians from the Concertación took notice. They could no longer do as they pleased in this issue area. These constraints opened the door to mediated effects in the shift from an institutional governance regime based on a weak decentralized agency under the presidency's political control to a Ministry of the Environment, where cabinet rank gave environmental issues more political weight. Mediated effects were stronger in non-traditional renewable energy policy where there was a shift in the share of those sources to the energy grid significantly beyond what had been previously legislated.
For Giugni (2007) each type of effect entailed a specific relationship of social movement organizations to the policy process that explains the closeness or distance of policy to the movement's policy positions. In the Chilean case, direct impacts were largely confined to agenda setting, which resulted in close correspondence of movement postures and governing political platforms. However, subsequent policy formulation was of the indirect type and the resulting policies habitually deviated substantially from that agenda. Mediated effects later afforded closer correspondence. It is the relative closeness or distance of policy to the movement's positions that is important here, not whether movements or political actors have primacy in the process. Since movements are engaging with the political process, the latter will per force play a significant or even preponderant role in the final policy outcome.

Explaining social movement policy effects
For Giugni, and others, the linkages of movements to the policy process, plus protest, explain the type of effect of social movements on policy, whether direct, indirect, and mediated. 1 However, what explains the prevalence of one type of effect over the other and, perhaps more interestingly, shifts from one type to another? This draws attention to the larger political, economic, and social context in which social movement organizations and activists seek to advance their policy goals. Employing a political economy framework, I argue that a change in the balance of power among domestic and international economic social, political, and institutional actors, in addition to the linkages of movements to the policy process and protest, accounts for shifts in types of movement effects over time.
To construct that balance of power within which the action of movements occurs I draw on a comparative political economy approach that examines economic, social, and political structures along with ideational factors to identify actors, their interests, and power resources. Economic and political structures and ideas define most of the state-institutional, party political, and social actors involved in the policy process and their policy stances. Those structures also define the distribution of power in a society. Political economic structure and the distribution of power it supports define what social movements interested in institutional and policy change are up against and what their possibilities are (Gourevitch, 1986).
What does Chile's political economy tell us about the distribution of power and its relationship to the environmental issue area? Chile's military government (1973)(1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)) established a free-market economy open to international trade and finance. It relies principally on mineral and agricultural commodity exports to drive economic growth. The state's functions were sharply reduced. Most public enterprises and services were privatized. 2 In short, the state and politics should keep their hands off the economy so as not to distort the functioning of the price system as the most efficient allocator of resources and engine of economic growth. These principles remain in effect to this day (Edwards & Cox Edwards, 1991).
Chile's economic model rendered business elites politically very powerful. They possess a great deal of structural power in that they are by far the principal source of employment and economic growth. They frequently threaten investment strikes when government proposes to regulate them or tax them or otherwise take actions that impinge on business' capacity to operate relatively unconstrained. That includes, stricter environmental regulation and state support for green technology, more labour friendly policy, improving social insurance or constraining private sector provision of education and health care. Moreover, business elites are highly organized in economic sector-specific associations and an encompassing peak association. They are proficient in public opinion formation, lobbying, and policy analysis (Silva, 1992(Silva, -1993. In democracy, business elite's power has additional roots in the political pact between the democratic opposition that led the transition from authoritarianism (beginning in 1983) and the supporters of the economic and social model implanted by the military dictatorship. That pact set strong limitations on reform initiatives after re-democratization in 1991 that persist until today (Boeninger, 1997;Siavelis, 2010;Tecklin et al., 2011). During Chile's transition to democracy (1986-1990) the democratic opposition formed a broad-based centre-left party coalition that eventually became the Concert of Parties for Democracy (CPD or Concertación). After re-democratization it ruled uninterrupted for 20 years (March 1991 to March 2011). Since then a conservative coalition has ruled twice with a CPD interregnum. The CPD's major parties were the Christian Democratic Party (PDC) in the centre and the Socialist Party (PS) and its offspin Party for Democracy (PPD) on the left, and the Social Democratic Radical Party (PRSD) on the moderate left. The democratic opposition steadfastly supported using government to moderate the many social and economic injustices produced by Chile's laissez faire system.
During re-democratization the military government was in firm control of political liberalization as spelled out in its own constitution of 1980. It's primary interest, aside from protection of military personnel from human rights prosecution, was the maintenance of the free-market model with a weak state in terms of regulatory capacity. The military government and business elites deeply mistrusted the democratic opposition, believing they intended to return Chile to the bad old days of import substitution and "socialist chaos." Given the military government's strength the opposition coalition assured the military regime they would not alter Chile's free-market foundation. If they became government they promised to only seek market affirming or market compatible means to increase the provision of public goods. This pact set the conditions for policy continuity with moderate change. More importantly, the CPD honoured this commitment in the interest of democratic stability when it became government in 1991. This included thwarting more ambitious reform initiatives within their ranks, such as early efforts to establish a ministry of the environment and to champion more labour friendly policy.
After re-democratization supporters of the free-market socio-economic model also counted with strong party political representation. This came in the form of two conservative political parties that formed an electorally strong conservative coalition that changes name for every presidential election. The larger of the two is a libertarian party of the dictatorship: the Independent Democratic Union (UDI) along with National Renovation (RN), the remnants of the historic right. For a variety of reasons, including institutional "tethers" to democracy featured in the dictatorship's 1980 constitution, the conservative coalition forms a powerful opposition bloc in the national legislature that fiercely protects the free-market economic regime established during the dictatorship, and the business interests that dominated it. They consistently water down or veto the CPD's mild reform bills (Siavelis, 2010).
In sum, from the moment the opposition bloc coalesced in the 1980s, and especially once it decided to contest the 1988 plebiscite on General Augusto Pinochet's presidency, the CPD and its governments focused on mild reforms to the market society model of the dictatorship. Many of these reforms concentrated on expanding state provision of public goods, especially in social and labour policy, and regulating issues where unfettered economic activity manifestly caused socially detrimental externalities. The environmental issue area fell into the last category. The strength of the conservative opposition and self-censorship forced CPD governments to compromise, thus many policy initiatives, which were usually mild to begin with, were even more watered down than originally conceived.
These general conditions and relationships, which remained more or less constant between 1991 and 2017, meant that institutional and policy change in democratic Chile was, per force, gradual. This was especially true in the environmental issue area given its low priority in relation to other issue areas. Most change was of a layered nature, meaning that new institutions or policies were added to others already in place (Mahoney & Thelen, 2009). A key mechanism in a number of issue areas involved the connection civil society organizations had established with the CPD during democratization, especially the more leftist PS and the PPD. It allowed social movement organizations with expertise in new issue areas like the environment and gender to have a direct effect in shaping CPD policy agendas. This was especially so during presidential elections where they contributed to writing governing platforms. 3 Implementation of that agenda was another story. Pro-market business forces, their conservative party political allies in the legislature opposed environmental NGO policy proposals because they had the potential to influence their accumulation possibilities (Ospina et al., 2015). Thus, together with consensus seeking, confrontation adverse CPD government officials and party leaders they ensured that subsequent policy formulation bore little resemblance to the proposals of the environmental movements and activists. This resulted in the environmental movement's indirect effect in the policy process, especially in the first three CPD governments.
Nevertheless, there have been shifts to mediated effects since 2010. If the general contour of the balance of power remained constant, what changed at the margins? First, international political-economic factors shifted. External pressure for improvement in environmental governance institutions was significant. Second, the internal balance of power in the CPD was changing as were the parties themselves. Third, mounting socio-environmental conflict and protests and an expanding cycle of mobilization in 2011 that included environmental movement protests against mega-projects, students, labour, and regional movements pressured politicians to craft policy reform with movement activists.

From National Commission to Environmental Ministry
The Chilean environmental movement consists of a network of activists and organizations that raise collective challenges to the environmental governance regime in sustained interaction with elites, opponents, and authorities. Their repertoire of contention includes protest and engagement with the policy process. The main national environmental movement organizations relevant for this paper are: CODEFF (Comité pro Defensa de la Flora y Fauna), Instituto de Ecología Política (IEP), Chile Sustentable, Fundación Terram, Corporación Chile Ambiente, and the Consejo de Defensa de la Patagonia. CODEFF focuses on nature conservation. In the 1990s it was active in protest campaigns against mega-hydroelectric dams, but since then it works more through policy networks on forest policy and nature conservation more generally. The IEP is, as the name suggests, a leading organization for policy analysis and activism in political ecology; as of the 2000s it focuses more on consultant work. Chile Sustentable has emerged in the 2000s as a leading policy advocacy and activist organization, currently it specializes in energy issues. Fundación Terram and Chile Ambiente specialize in policy analysis and consultancy work (Mattila, 2005). All of these organizations regularly support protest campaigns. The Patagonia Defense Council is an example of most of these organizations working together on one multi-year national campaign to stop a complex of hydroelectric dams in Chilean Patagonia. Several of those organizations have developed expertise in water rights and energy policy. These organizations tend to be supportive of each other, preferring to keep a common front for the public given that they face such an uphill battle in the environmental issue area (Mattila, 2005). All advocated a strong environmental ministry.
The CPD championed socio-economic and cultural issues ignored by the military dictatorship, including the environment. Thus, as a constituency of the CPD, and because of their technical expertise, the leadership and activists of the environmental movement organizations participate in the CPD's policy networks; and, thus, are able to directly influence policy agendas by detailed drafting of their policy platforms. They also work with international organizations, such as the United Nations and the OECD. However, Chilean environmental organizations also engage in contentious action to pressure authorities, including public relations campaigns and classic demonstrations in conjunction with lobbying. These methods are generally used to critique straying by politicians from the agreed upon environmental policy agenda and to publicize alternatives hoping to be included in the policy process to ensure closer correspondence between their proposals and policy outcomes -in other words they aspire to achieve mediated policy effects. Within this mix of tactics, public protests have played a significant role in anti-mega project campaigns in Chile.
The analysis that follows traces the policy process in the development of environmental governance institutions in Chile since the transition to democracy in 1990. The establishment of a strong environmental ministry was a key institutional demand of environmental movement organizations from the beginning. Movement organizations were able to get that demand on the CPD's policy agenda. However, due to the political pact struck during the transition to democracy they only had an indirect in policy formulation during three consecutive CPD governments. Those governments formulated policy without their participation and created a much weaker environmental governance regime without an environmental ministry. I also show how mounting socio-environmental conflict, largely around anti-mega project protests, kept the movements agenda alive. I then show how changes in the CPD's internal balance of power coupled with a shift in international conditions altered the overall balance of power set in the democratization pact sufficiently to create a ministry of the environment.

Democratization, political pacts, and movement policy impacts, 1989-2005
Beginning with the transition to democracy in 1990, environmental movement organizations such as CODEFF and Instituto de Ecología Política proposed the establishment of a strong institutional regime for the environmental issue area anchored in a Ministry for the Environment. In their view, cabinet rank would place the environment on a more equal footing with economic sectors it would regulate. They placed this proposal on the policy agenda of the CPD's very first presidential campaign in 1989 led by Patricio Aylwin, who went on to win the election in December of the same year. They basically wrote the environmental platform for his campaign. Thus, the proposal for a strong environmental institutional regime was a direct effect of the environmental movement organizations' engagement, indeed collaboration, with the CPD (Silva, 1996(Silva, -1997.
Aylwin and the CPD won the 1990 presidential election in which democracy returned to Chile. Initially, environmental movement organizations had mediated effects in the formulation of a framework environmental law; thus, there was an opportunity for closer correspondence between the movement's agenda and actual policy. Environmental movement organizations had close connections to the PS and the PPD; and it was these parties whom the Christian Democratic presidency put in charge of environmental policymaking. Consequently, environmental activists collaborated with the PS/PPD and the Ministerio de Bienes Nacionales (which managed public land and properties) in drafting an environmental framework bill. Initial drafts took a progressive thrust, including a proposal for a ministry of the environment (Silva, 1996(Silva, -1997.
However, conflicts between environmentalists in Bienes Nacionales and the presidency ensued. The presidency side-lined the environmental movement from the policy process and worked with a business oriented environmental think tank. With that, movement effects in policy formulation shifted from mediated to indirect. The movements had placed the issue on the policy agenda but then government made environmental policy without them substantially changing the content of policy away from what the movements had proposed (Silva, 1996(Silva, -1997.
Why did this happen? Here we see the balance of power between proponents of a strong vs. weak environmental governance regime at work. The presidency, whose main priority was to oversee a smooth transition to democracy from the military regime, worried that the progressive agenda of environmentalists in Bienes Nacionales clashed with promises made to the military and business elites that the basic socio-economic model implanted by the dictatorship would not be altered. The conflict centred on a proposal for a strong state institution with significant oversight and regulatory capacity over the private sector. As a result, the presidency took control of environmental policymaking away from the PS/PPD and Bienes Nacionales. Under the presidency's close tutelage, policymakers created a National Commission for the Environment (CONAMA) that was directly under the presidency and charged it with drafting an environmental framework law. CONAMA did not involve environmental movement organizations to work with it (Silva, 1996(Silva, -1997. The environmental framework law took a very different direction than that advocated by the environmental movement. The CONAMA became the principal environmental governance institution. It was decentralized, inter-ministerial, without cabinet rank, and had weak regulatory capacity. Environmental impact reports (EIRs) required for new large-scale development projects were one of its main regulatory instruments. The EIR process offered the only institutionalized opportunity for civil society commentary on development projects (Silva, 1996(Silva, -1997. Regulations to Environmental Impact Assessment System (SEIA) were market affirming and business friendly. The process unfolded without social movement organization input and was not completed until 1997 in a second CPD government headed by Eduardo Frei, Jr. also a Christian Democrat (1994-2000) (Pizarro & Vasconi, 2004).
From this point on, reform of this weak, business-friendly institutional regime became an important goal for Chilean environmental movement organizations. Achieving that goal required keeping the demand for a stronger, more autonomous institutional regime on the political agenda and finding propitious political circumstances to realize that objective. Mounting socio-environmental conflicts beginning in Aylwin's administration and then snowballing in Frei's and into Lagos' government and beyond played an important role in that quest. Some were prolonged, highly visible national campaigns that combined protest with legal action to halt new mega-projects and to create favourable public opinion for institutional change. The fact that demobilization characterized most social movement-government relations during this period added to their visibility. We now examine socio-environmental conflicts and policy impacts from 1994 to 2005. Three protracted, high visibility national campaigns laid-bare what environmental movement organizations considered to be the inadequacies of Chile's environmental governance regime (Delamaza et al., 2017). The first mega project involved a series of hydroelectric plants to be constructed on the upper Bío-Bío River in south-central Chile. The project officially began in the last year of the dictatorship (1989) when ENDESA, a Spanish transnational energy corporation, in association with Chilean partners, proposed two major centrals: Pangue and Ralco. The project then wound its way through the permitting process during the Aywin and Frei administrations (Silva and Rodrigo, 2010). Environmentalists began campaigning to stop their permitting and construction during Aylwin's government on grounds that the Pangue mega-dam and reservoir and the Ralco central threatened a pristine ecosystem and the territory of the semi-nomadic Pehuenche Mapuche people. Activists from major national environmental organizations, including CODEFF and the Instituto de Ecología Política, coalesced in the Grupo de Acción del Alto Bío-Bío (GABB). 4 The GABB mounted protests, media campaigns, and engaged CONAMA functionaries to persuade it to reject the project's EIR.
Their efforts failed given the unfavourable balance of power embedded in the pact for the transition to democracy. CONAMA approved the EIR, loans were granted, and the Pangue Dam was inaugurated in 1997 during the Frei administration (Baquedano, 2004). Frei himself unequivocally backed the project stating, "Nothing will stand in the way of the development of this country" (Pizarro & Vasconi, 2004, 5). Indeed, the central government suppressed an unfavourable CONAMA technical report allowing the EIR's approval. GABB director Juan Pablo Orrego, after receiving distraught apologetic calls from CONAMA functionaries saying they had been forced to approve Ralco over the technical department's recommendations, stated "CONAMA gave way to pressure from the government" (Figueroa Fagandini, 2011). GABB then mounted a second national campaign to stop construction of the Ralco central. It organized highly publicized marches and demonstrations in the capital city (Santiago) and in the south central region with students, women's organizations, indigenous people's organizations and others. Media campaigns further publicized the conflict, which dragged into the presidency of Ricardo Lagos, when the project was finally approved (Silva and Rodrigo, 2010).
Pascua Lama was a second major national socio-environmental conflict that erupted in Frei's administration and dragged on to vex others. In 1996 a consortium led by the international Barrick Gold Mining Company proposed a mega project to extract gold, silver and copper at 4,000 meters altitude in Northern Chile (Atacama region) and Argentina (San Juan Province). The major investors in the project immediately lobbied the Chilean and Argentine governments for a bi-national treaty permitting the project, which was signed in 1997 with the protocol that complemented it following in 1999 (INDH 2012, 62-64). The mining project threatened two glaciers that were the watershed for downstream agricultural communities, among them Diaguita peoples. National and local environmental organizations, alongside local community civic groups and municipal officials publicized these highly negative socio-environmental impacts and lobbied CONAMA to reject the project as proposed. Their efforts fell on deaf ears. (Silva & Rodrigo, 2010;INDH 2012, 62). 5 A third mega-project conflict that plagued the Frei and Lagos administrations began in 1994 and involved the permitting, construction and operation of a pulp and paper plant by Celulosa Arauco y Constitución (CELCO) near Valdivia in Southern Chile. CELCO was the flagship firm of a major Chilean conglomerate and the environmentalists' main concern was pollution from effluents and the siting of their catchment. If effluents spilled into the ocean they would affect fisheries upon which local artisanal fishermen depended for their livelihood. If effluents were directed to the Anwandter Wetlands, a protected area, they would threaten the local fauna, especially black-necked swans (Silva and Rodrigo 2010;Instituto Nacional de Derechos Humanos [INDH] 2012, 244). As in the Pangue-RALCO case, environmental movement organizations and community activists mounted a spirited, locally and nationally visible campaign. Local fishermen in Mehuín prevented CELCO engineers from carrying out studies necessary for their effluent pipeline EIR, an act publicized by the national media. However, ignoring environmentalists, CONAMA approved a pipeline to dump effluents into the Anwandter Wetlands Refuge despite legal irregularities and incomplete EIRs (Silva and Rodrigo, 2010;INDH 2012, 244).
The Lagos government proved no different than those before; it approved practically all of the EIRs for new projects and ignored institutional reform. Ultimately it bore the political consequences when the CELCO case pressed itself on the national consciousness. Illegal discharges of toxic effluents raised the level of pollution in the Anwandter Wildlife Refuge to such an extent that in 2004 thousands of treasured black-necked swans died or migrated, along with fish, flora and fauna (INDH 2012, 244-247). The scale of the disaster in a popular recreational area was shocking to the public. Prime time television coverage turned the event into a national scandal (OLCA 2004). The case highlighted irresponsible corporate behaviour coupled with permissive, lax government regulation. Business could clearly not govern itself (González, 2005). The tide of public opinion was turning in favour of reform of the environmental governance regime. For example, in 2005 a national opinion poll showed CONAMA's credibility below 20 percent (CERC 2005). Environmental organizations capitalized on these events to press their critique of the environmental governance regime (Fundación Terram 2005).They protested, mounted media campaigns, and worked within the environmental governance system challenging EIR approvals and suing for legal injunctions through the courts.
The upshot of these events was threefold. First, intensifying socio-environmental conflict and high profile environmental disasters raised the public's perception of environmental problems and by 2004 unequivocally generated public pressure for government to do something about them. Second, social movement organizations sharpened their critique of Chile's environmental governance regime on the basis of ample evidence. Third, environmental movements refined corrective policy proposals and lobbied both politicians and the think tanks that crafted their policy positions. Thus, environmental movement organizations, like Fundación Terram, argued that socio-environmental conflicts exposed CONAMA's weakness, its lack of authority and autonomy. It was subordinate to the presidency and run by the interests of line ministries. Environmentalists claimed that these characteristics turned EIR evaluations into a closed political, and therefore arbitrary, process that generally protected business interests over the social and environmental dimensions of sustainable development. They contended that the Environmental Framework Law, and the resulting environmental governance regime, revealed that CONAMA could not handle the contentious politics that broke out. Its lack of ministerial standing and control over line ministries and business interests cost two of the CONAMA presidents their posts in Frei's government. The socio-environmental conflicts revealed the political, rather than legal, procedural, and technical, rationale for deciding compliance with environmental standards and laws, including protected areas and protected peoples (Pizarro & Vasconi, 2004).
In 1999, the fact that expanding socio-environmental conflicts had gained public visibility and generated unfavourable public opinion for the CPD probably affected Socialist CPD presidential candidate Ricardo Lagos' decision to feature environmental issues in his platform. It was a signal that under his party's leadership the CPD would take a more progressive direction than under the Christian Democrats. His campaign worked closely with a working group of environmental movement organizations, activists and academics calling themselves "Comando Verde Por Lagos," that wrote the candidate's environmental platform in May 1999 (Comando Verde por Lagos, 1999).
Once again, given their connections to the PS/PPD a group of environmentalists had a direct effect on generating an environmental policy agenda for a CPD government. That agenda addressed the central points of the environmentalists' critique of the environmental governance regime. Moreover, these were policy recommendations that the environmental movement had advocated since Aylwin's government. They included (1) establishing a strong state authority for the environment, (2) perfecting the environmental impact reporting system, (3) promote green materials management production, (4) design and implement land use planning protocols, (5) protect natural resources, (6) deepen citizen participation in the SEIA, (7) actively implement international environmental treaty obligations, (8) redesign Santiago's public transportation system to help decontaminate the metropolitan region.
However, Lagos' government disappointed environmentalists. As with previous CPD administrations, it did not implement the agreed-upon platform's planks, with the exception of metropolitan transport. Socio-environmental conflicts were generally resolved in favour of business. Projects continued forward, EIRs were approved, polluting plants operated without even end-of pipe pollution control technology, much less green technological innovation (INDH 2012). In this context, the Ralco, Pascua Lama, and CELCO campaigns were, among other objectives like stopping the projects themselves, a means to keep environmental problems and the role of environmental institutions in them in the public eye as a means to keep environmental regime reform on the policy agenda. By 2004 public opinion had decisively turned in favour of the environmental activists position against a CPD governing coalition perceived to be complacent. A string of other highly visible environmental disasters, like contamination from wastewater treatment plants that fouled densely populated metropolitan Santiago (INDH 2012, 176-183), added to that perception.
Building on the evidence of these events for environmental organizations like Fundación Terram, Chile Sustentable, and CODEFF honed their perennial policy platform for the 2005 presidential elections to drive home the point that current institutions were ineffective. Thus, they agreed on a platform that included the following points: (1) A strong ministry of the environment with an autonomous oversight agency to ensure regulatory compliance; (2) Strengthen the Environmental Impact Assessment System's technical capabilities and insulate it from political manipulation; (3) Expand citizen participation in environmental governance; (4) Strengthen institutional capacity in nature conservation and biodiversity protection, beginning with the creation of an agency devoted to the issue; (5) Develop land-planning protocols; (6) Increase regulation and prevention of urban pollution, especially air quality; (7) Strengthen regulation to mitigate the socio-environmental externalities of natural resource extraction; and (8) Develop non-traditional renewable energy sources (Pizarro & Vasconi, 2004).

Environmental governance reform, 2005-2010
Mounting socio-environmental conflict and the growing clamour from public opinion enabled environmentalists to keep pressing for stronger environmental governance institutions. By 2005 that pressure was sufficiently strong to put environmental governance reform on the agenda of all of the candidates for president, not just the CPD's. Michelle Bachelet of the PS became the CPD's candidate and won the 2005 presidential elections on a platform of greater inclusiveness. Among other measures, her presidency undertook an environmental governance reform that established a ministry of the environment, a long-standing demand of the environmental movement.
Once again environmental activists' connections to political parties facilitated their capacity to directly influence policy agendas. However, they were dependent on political actors, primarily the presidency but also increasingly in the legislature, for policy initiation and for their participate in policy formulation. This required changes in the balance of power of the democratic transition. In short, environmental movements depended on the political will of the executive for their involvement in policy formulation but that will, in turn, depended on the dynamics of the balance of power established during the transition to democracy and on external factors. I argue that two relatively minor changes in the conditions of that balance of power influenced this more favourable outcome. One was a change in presidential candidate selection in the PS and the CPD. It permitted Bachelet, an outsider to the PS party machine, to become the CPD's presidential candidate and eventually president. Thus, the weight within the CPD shifted further left with a more progressive reform-minded and slightly more autonomous candidate. The second factor involved Chile's position in the international political economy. It was up for OECD accession. One of the conditions, albeit of lower importance in relation to other economic reforms, was that Chile should modify its environmental governance regime to conform to OECD member standards.
In the end, these two factors brought environmental institutional reform closer to movement policy positions, but not as close as movement organizations hoped. Initially they were poised to generate mediated effects in the legislature that would have brought the bill much closer to the environmental movement's positions. However, the overarching conditions of Chile's transition pact balance of power allowed conservative pro-business opposition to side-line environmentalists and significantly water down institutional reform. Thus, more customary movement indirect effects prevailed. We now examine how environmental movements contributed to shape Bachelet's policy agenda.
There was a great deal of coincidence with respect to institutional reforms between the policy proposals of insider environmental organizations, the parties of the CPD, and Bachelet's presidential campaign. They all proposed a ministry of the environment, superintendence to administer natural resources, conservation and biodiversity, and stricter enforcement of environmental laws and regulation. There was also coincidence in proposing increased citizen participation in environmental impact evaluations, especially by extending the time for citizen review and making information more complete, timely, and transparent (Larraín & Schauenburg, 2006).
How did social movement organizations shape the agenda of Bachelet's governing program, especially since she had own advisory think tank, Expansiva? First, social movement organizations, like Fundación Terram, gave presentations to Expansiva (Pizarro, 2005). Secondly, the political parties of the CPD, to whom Bachelet was an outsider, had developed some independent technical capacity to analyse environmental issues. Thus, Terram, Chile Sustentable, and Chile Ambiente, among others, advised the environmental committees of CDP political parties, especially the PS/PPD. They also sent detailed policy position papers and gave presentations to other CPD parties, such as the PDC (which drafted a document specifically intended as an input for Bachelet's governing program). The programmatic positions of these parties on change in environmental governance also coincided with the environmental movement's positions (Larraín & Schauenburg, 2006).
The timing of specific proposals as a condition of environmental movement support and the presidential campaign's release of its environmental governing platform also support the claim that movement organizations influenced the CPD's environmental agenda. During 2005 environmentalists prepared an "Agenda Ambiental de ONGs Ecologistas y Ambientalistas: Prioridades para el período 2006-2010." They distributed this document to all of the presidential candidates in September 2005. Michelle Bachelet released her program "Nueva Política Ambiental," closely aligned with the movements' proposals, on 15 October. Bachelet actively courted and received the endorsement of environmental organizations by signing the 10-point Acuerdo de Chagual with them on 21 November 2005 (Larraín & Schauenburg, 2006).
Another factor that raised the relevance of environmental issues for policymakers was the fact that Chile under Lagos had applied for admission to the OECD. In May 2005 the OECD officially released an evaluation of Chile's environmental regime (OECD-ECLAC, 2005). The OECD report's critiques and recommendations were very much in line with those that the environmental organizations had been making since the early 1990s. Indeed, those organizations influenced the OECD report's content. They had input in the OECD's information gathering process. ECLAC (UN Economic Commission for Latin America and the Caribbean) was the OECD's local institutional counterpart and responsible for information gathering and analysis. The heads of ECLAC's environmental department shared the environmental organizations' positions; indeed they were part of broader policy network that included activists from the principal environmental movement organizations. Environmental organizations on occasion also gave formal presentations to the OECD fact-finding group (Fundación Terram, 2004).
Thus, although the OECD report was an important factor in shaping Bachelet's campaign's environmental agenda it was not decisive (Fundación Terram, 2009Terram, & 2008. Nevertheless, her objective that Chile be admitted to the OECD was a significant factor in securing implementation of that agenda after Bachelet won the presidential election. 6 Still, it was domestic politics that shaped the substance of the institutional reforms that emerged from the policy formulation process. I now turn to an account of that process of moving from agenda to institutional reform and the role of social movement organizations.
The Bachelet administration introduced an environmental governance regime reform bill to the legislature in July 2008. The law passed in January 2010 just before the end of her presidency. It established a Ministry of the Environment, Superintendence for the Environment, and a System of Environmental Impact Assessment. It also mandated the establishment of Environmental tribunals. It left the creation of a Conservation and Biodiversity Superintendence for future legislation. The CONAMA -Council of Ministers -would continue to function in collaboration with the Ministry of Environment.
However, the final law mandated a significantly weaker environmental governance regime than was the preferred environmental movements. This occurred because after some innovative political manoeuvring in the end movement organizations only had an indirect effect in policy formulation. Having placed the issues on the agenda, ultimately, they were frozen out of the legislative process (Fundación Terram, 2009). This is explained by the relationship of the movements to political parties and of how the post-dictatorship balance of power influenced the relationship of the executive branch to the legislature. In other words, political actors drove the policy formulation process, but the presence or absence of environmental movement organizations conditioned the substance of policy outcomes.
Policy formulation unfolded as follows. As was traditional in Chile, policy initiation rested with the presidency. Bachelet's administration formulated a bill and submitted it to the legislature in 2008 without open discussion or civil society participation. In other words, environmental organizations were, once again, relegated to an indirect effect in the process. Thus, important features of the bill were not congruent with the environmental organizations' proposals (Fundación Terram, 2009). It offered a substantially weaker environmental governance regime.
The executive then submitted the bill to Deputies, the lower chamber of Congress. Now that there was an actual bill in the legislature, environmental organizations had an important political arena opened up to them. Moreover, the legislative elections of December 2005 shifted the balance of power in favour of more progressive policies. Voters, tired of CPD establishment representatives that blocked reforms to the socio-economic system imposed by the dictatorship and its inequalities, had elected more progressive ones. Thus, given their connections to political parties, and hence to legislators in the lower chamber, activists had good access to the process as advisors to deputies in the Environmental Commission (Biblioteca del Congreso Nacional de Chile [BCN], 2010). Many of their proposals for a stronger governance regime were written into the bill, which Deputies approved. Thus, in the lower chamber environmental organizations had a mediated effect on policy.
However, the institutional embedding of Chile's post-authoritarian balance of power in the Senate negated that effect. The Senate -where conservative proeconomic growth and low regulation forces dominated -had the last word in the legislative process. Its environmental committee had made many critical and contradictory observations to the bill from Deputies (Fundación Terram, 2009). Fearing that the bill would get bogged down in committee, the Senate Committee for the Environment drew up a protocol to expedite discussion. This meant that discussion of points of conflict with conservative senators involved quick and dirty compromises that gutted the more aggressive positions environmentalists and their supporters in Deputies had made (Fundación Terram, 2009). Hence, in the final analysis, once again, movement activists were frozen out and only had an indirect effect on the policy outcome. Thus, the law created a weaker environmental governance regime with diminished regulatory and oversight authority.

Energy policy: crafting the Law for Non-Traditional Renewable Energy
The contrast of the environmental regime issue area and the energy sector is interesting for several reasons. First, it highlights important differences in the political economy of the policy process. Whereas, the former affected the accumulation possibilities of all productive sectors of the economy (with spill over into finance), the latter primarily affects the energy-producing subsector. Hence, broad cohesive opposition tends to be more durable in the former than in the latter. Second, external factors, such as OECD accession, played an important role in environmental regime reform, whereas in the energy sector politics were more national in scope. Third, movement activists had learned some lessons from the struggle for institutional reform. The discussion opens with an overview of the energy sector. It then analyses the policy process in two steps. It gives an account of events in which the policy process follows a dynamic similar to environmental regime reform. It concludes with an account of divergence from that pattern in Sebastián Piñera's administration (2010)(2011)(2012)(2013)(2014).

Movement policy impacts
Among the many issues that environmental movements had been advocating since redemocractization was a shift to greater use of non-traditional renewable energy sources. Chile had a market-based approach to energy dating back to the 1982 Energy Law promulgated by decree during the military government and relied on imported natural gas from Argentina and hydropower (Maillet & Rozas, 2017). Two shortages shook the foundations of this arrangement: the drought of 1998-1999 and the Argentine gas crisis of 2002-2004. In the aftermath, energy prices soared making Chile's power among the most expensive in the world (BBVA, 2014). Thus, energy security became a policy issue.
In the first round of energy policy reform the debate turned on expanding thermal energy (dirty coal), hydropower from regions not subject to drought, importation of natural gas from other countries, nuclear, and non-traditional renewable sources. There was an upsurge of thermal energy production projects during Bachelet's first administration (2006Bachelet's first administration ( -2010. But in an effort to keep commitments to the environmental movement her government also pledged to expand non-traditional renewable sources. Environmental organizations wanted an increase to 20-30 per cent of energy generation by 2025. However, in April 2007, the executive introduced a bill that only limited increases to 10 percent by 2020. The "10/20" bill sailed through the legislature and became law in March 2008 (BCN, 2008).
What impact did environmental organizations such as Chile Sustentable, Fundación Terram, and Chile Ambiente have on the policy process that ended in the 10/20 Law? Using their well-established connections to CPD political parties, they put non-traditional renewable sources on the policy agenda of Bachelet's governing platform in 2005 (Larraín & Schauenburg, 2006;Bachelet, 2005). But the movement's proposal was a political hot potato given that Chile's energy companies opposed such initiatives. The unfavourable balance of power towards environmental issues caused Bachelet's government to table the matter.
However, a new political actor took up the agenda. In July 2006, in frustration with government's lack of interest, a group of progressive representatives in legislature's House of Deputies introduced a bill on non-renewable energy sources. Environmental organizations in their capacity as advisors to key parliamentarians helped formulate the bill, which set a target of 30 percent non-traditional renewable energy by 2025 (BCN, 2008). As previously discussed, the lower chamber's increased activism responded to the influx of more progressive deputies as an expression of voter impatience with the pace of reforms to inequalities created by the socio-economic system implanted by the military regime. This change afforded environmental activists a space to pursue mediated policy effects and, thus a chance for closer congruence between environmental activists' policy positions and the proposed bill.
Mindful of the political settlement that supported democracy in Chile and the balance of power that underlay it, the executive, stung into action by the "radical" energy bill, countered by sending its own less demanding bill in April 2007. The administration kept CPD party discipline and environmental movements had little if any direct impact on it. They testified in key legislative hearings, explaining that a more comprehensive law should aim for greater percentages, establish adequate funding as incentives to production in part funded by traditional energy companies through a surcharge on their energy, state support for the promotion of non-traditional renewable sources, and penalties with noncompliance to targets (BCN, 2008). None of these recommendations made it into the law promulgated in March 2008. This, then, was a classic indirect effect of movement organization contentious politics.
This outcome set up round two of energy policy reform. During Bachelet's first government, thermal energy from Chile's ample dirty coalfields became the administration's first choice to address the energy crisis. Numerous projects were introduced into the EIR system between 2006-2010. Many of them became sources of socio-environmental conflict. Indeed, compared to the 1990-2005 period, they were the major source of conflict (INDH, 2012). The environmental movement was disturbed by this turn and criticized the government for it. Adding insult to injury, the administration also supported a mega hydroelectric project in Chilean Patagonia. Several rivers would be dammed and energy would be brought to the central grid via a yet-to-be-built several thousand km long high-tension transmission line. Bachelet's government initially backed the project, dubbed HidroAysén (Silva, 2016).
HydroAysén became the centre of the largest, national environmental protest campaign of the era in 2011, culminating in the rejection of the project in 2014 (Silva, 2016). It was led by the Patagonia Defence Council (PDC), which included Chile Ambiente, Chile Sustentable, Instituto de Ecología Política, CODEFF, and Fundación Terram among others. Beginning in 2007-2008 the PDC conducted a smart media campaign that turned public opinion against Hy-droAysén and started the discussion on alternative renewable energy. For example, in 2008 the PDC commissioned a poll by a reputable firm in which 53 percent of respondents opposed the project, whereas before the media campaign took off in 2007 54 per cent had favoured the project. In April 2011, during the centre right administration of Sebastián Piñera a new poll indicated 61 per cent disapproval rating for HydroAysén (Consejo de Defensa de la Patagonia Chilena, 2014).
Meanwhile, the protest campaign took off in May 2011 after Piñera's government approved HydroAysén's EIR. A massive demonstration (especially for an environmental issue) of 30,000 to 40,000 people rocked the government. Environmentalists were quickly joined by student protests against privatized education, public sector labour unions against flexibilization of labour, and a regional movement in Aysén for government assistance for a special region. The various movements then began coordinating mobilization in a Mesa Social for a New Chile in 2012. Hundreds of thousands of people were demonstrating across Chile in a cycle of mobilization not seen since the days of political polarization in the early 1970s (Donoso & von Bülow, 2017).
As in the past, then, mounting socio-environmental conflict kept issues in the public eye generating public demand to do something. Doing something, in this case, aside from reforming the environmental governance regime, involved generating cleaner, cheaper energy. Social movement organizations' contentious politics, and protest specifically, certainly played a role in raising the policy profile of the issue. But social movement organizations, such as Chile Sustentable, Fundación Terram, and Chile Ambiente also worked their networks with legislative representatives. They advised senators and deputies on the environmental committees who were looking for more aggressive policy than what the executive offered. In classic mediated effect style, movements and protest placed or kept alive issues on the policy agenda and the worked with politicians who shared their concerns (BCN, 2014).
The result was that the Senate Committee for the Environment introduced a bill in September 2010 that contained many of the points whose absence the environmental movement critiqued in the executive's sponsored law of 2008. Environmental movement organizations helped to formulate the initial bill, which called for 30 per cent non-traditional renewable energy inputs by 2025. The process dragged on for three years and the law, mandating 20 per cent by 2025, was passed in October 2013 (BCN, 2014).

Explaining outcomes
The above account initially reveals the same pattern we saw in the environmental regime reform process. Movements had significant direct impact on CPD agenda setting. Environmental movement linkages to the CPD ensured that Bachelet's first government (2006Bachelet's first government ( -2010 was open to movement proposals for a law mandating percentage targets of non-traditional renewable energy sources in the national grid. Given the balance of power inherent in the political settlement struck at the founding of Chilean democracy, as manifested in the politics of the Senate, the Bachelet administration then froze movements out of the policy formulation process ensuring a watered law more favourable to established business interests. Thus, movements only had indirect effects in the formulation of the first law to promote non-traditional renewable energy.
However, the political economy of the energy sector differed from that of environmental regime reform and movement activists had learned from the past. The energy sector was a more circumscribed policy issue area. The effects were more concentrated on communities; targets were more easily identified and plentiful. This had the consequence that activists could think of mobilizing sustained protests against them. However, doing so involved a learning process by activists. For the most part, they had relied on insider tactics -working within institutions -to affect policy. By the end of the first decade of the twenty-first century they had concluded that was ineffective and that organizing protest campaigns to pressure the political establishment was necessary.
Protest, it turned out, was effective, especially since in 2011 it snowballed into a sustained cycle of mobilization involving environmentalists, students, indigenous peoples, and, later regional movements and public sector workers. The cycle of mobilization influenced a change in the balance of political power between the business sector and environmentalists in the legislature. The Senate, the more important of the two chambers, now sided with environmentalists. In part this was because the 2009 parliamentary elections led to the entry of more environmentally minded independents in the conservative camp and a business lobby for non-traditional renewals had formed. This shift opened up a new source of movement influence. More progressive deputies and senators used the pressure of the cycle of mobilization to include movements in policy formulation. Thus, we see a shift from indirect to direct and mediated policy effects of social movements. The latter facilitated greater congruence between the movements' policy stances and the resulting legislation.

Conclusion
When analysing the policy consequences of movements researchers are tempted to determine whether social movement or institutional political actors -frequently conceptualized as political will -drive policy change. In my view, this misses the point. The key question is more by what mechanisms, to what extent, and under which circumstances do social movements have an impact on the various stages of the policy process? In short, how can we theorize the complex relationship between social movements, policies and institutions in democracies? An underlying assumption is that movement organizations engage in institutionalized politics. The key question is: from a position of relative powerlessness how can movement activists place issues on policy agendas and influence the subsequent policy process?
The literature in political sociology is useful because it seeks to establish the mechanisms by which they do so. The Chilean case supports the proposition that the linkages of social movement activists to political parties and policy networks are an important causal mechanism. I would add that for the policy proposals of movements to gain traction those parties and contacts in the executive branch must have political clout. From here, and building on Giugni (2007), one can establish whether movement organizations have direct, mediated, or indirect policy impacts. This allows us to establish the degree of correspondence of policy outcomes to the movement's policy declared preferences. Thus, "political will" may often drive policy change but which type of impact prevails matters for the substance of policy and/or institutional reform.
A policy process approach opens analysis to the different stages of the process in which social movements may have these impacts. This offers analysts a tool to be more precise about which impacts are operating when. For example, the Chilean case showed that movement organizations could have direct impacts in agenda setting but usually only indirect effects in policy formulation, and occasionally mediated effects. The stages approach to the policy process also allows us to be more precise about the impact of protest and public opinion on policy outcomes. It can have effects in agenda setting, as is well known, but also in keeping pressure on politicians during policy formulation, even when movement organizations are participating in them. It can even play a role in policy feedback effects as they make public the shortcomings of existing policies. However, these linkages and effects do not take place in a vacuum. Movements operate within an overarching balance of power of social and economic forces that is reflected in political institutions, such as political party systems, the state, and courts. These political institutions, however, are not monolithic or unchanging. Hence they can offer spaces for movements to advance their policy agendas.
This paper applied a political economy approach that examines economic, social, and political structures along with ideational factors to identify actors, their interests, and power resources as a framework to analyse the complexity of the balance of power within which social movements act (Gourevitch, 1986;Di John & Putzel, 2009;Hickey, 2013). Economic and political structures and ideas define most of the state-institutional, party political, and social actors involved in the policy process and their policy stances. Those structures also define the distribution of power in a society. Political economic structure and the distribution of power it supports define what social movements interested in institutional and policy change are up against and what their possibilities are.
It follows that changes, even marginal ones, in those structural and cognitive factors will affect the distribution of power among economic, social, and political actors and the possibilities for policy and gradual institutional change. Social movements, their organizations and activists, may help generate those changes through protest and the mobilization of public opinion. They may also exploit marginal or major changes in institutional structures that undergird an overarching balance of power caused by shifts in voter preferences, economic problems, or external pressure. Indeed, the literature has established that agenda-setting effects are the most usual result. However, once past the agenda-setting stage they will at best, if at all, be one more actor in the policy process, and although rarely the most central one, nonetheless a significant one.