Institutional thickening and innovation: reflections on the remapping of the Great Bear Rainforest

As a response to forest conflict, contemporary remapping refers to re-evaluations of resource values, new and diverse forms of governance among stakeholders, and compromises within patterns of land use that give greater emphasis to environmental and cultural priorities. This paper elaborates the processes of remapping by examining the role of institutional innovation in conflict resolution, with particular reference to the iconic Great Bear Rainforest of British Columbia. After years of conflict and protest, peace in the Great Bear Rainforest was heralded by an interim agreement in 2006, with final ratification likely in 2016. Conceptually, a four-legged stakeholder model identifies the main institutional interests and their interactions through learning and bargaining. New forms of governance were created to bring the stakeholders together in constructive dialogue and then to reach and implement acceptable bargains. Analytically, the paper examines how this agreement has worked in practice by reflecting on the emergence of novel institutions that integrate the interests of key stakeholders. The discussion identifies six bilateral negotiations between: industrial and environmental interests; federal and provincial governments and aboriginal peoples; government and environmental interests; government and industry; industry and aboriginal peoples; and environmental groups and local communities. The remapping process has produced a thickening architecture of institutions that remain experimental even as they seek to promote sustainability, resilience and legitimacy.


Introduction
Forest conflicts have a long history (Widick 2009) and have become a widespread feature of contemporary globalisation. Broadly defined as disagreements and disputes regarding access and management of resources, incompatible activities that seek to restrict one another and clashes among diverse institutional interests over the control and use of resources (Gritten et al. 2013;Hayter et al. 2003), forest conflicts feature a wide-ranging mix of actors and motivations. 'Wars in the woods', a metaphor with sometimes literal connotations, are typically intense, long-lasting and locally contingent. Yet a significant impulse driving forest conflicts around the world is opposition by diverse environmental, aboriginal and related communities to the vested interests of forest commodification in favour of the non-industrial, sustainable values of forests, and the emergence of these stakeholders forces the remapping of the forest and its human relationships.
As an umbrella term, remapping refers to two closely related processes in support of new forms of sustainable and equitable resource governance: the revaluation of resources from a commodity base to incorporate environmental and non-consumptive uses, and specific changes to land use designations that derive from these revaluations (Hayter 2003;Clapp 2004;Affolderbach et al. 2012; see also Brogden and Greenberg (2003) on re-territorialisation). Remapping serves as both normative metaphor for the conflicts and prescription for new regional plans and forms of governance that generate new institutions to address stakeholder interests in ways that replace conflict with cooperation. The contested creation of the Great Bear Rainforest on the central coast of British Columbia, Canada, provides an iconic case.
The Great Bear Rainforest (GBR) stretches from Bute Inlet to the Alaskan border and is 6.4 million hectares in extent, comprising about one-quarter of the world's remaining old growth coastal temperate rainforest, and a rich range of flora and fauna (Figure 1;McAllister et al. 1997). This large, remote and sparsely populated territory contains just over 22 000 people, half of whom identify as aboriginal and belong to 27 nations. With roots in BC's 'wars in the woods' that flared in the early 1980s, the region became the focus of intense conflict that culminated in a 2004 stakeholder agreement (CCLRMP 2004) in which onethird of the region was protected, logging companies committed themselves to ecosystem-based management in the working forest and commitments were made to the development aspirations of indigenous peoples. This agreement was ratified in 2006 by government and aboriginal peoples (called 'First Nations' in Canada), and painstakingly elaborated during subsequent years; the final legislated agreement was announced on 1 February 2016.
Although not without its critics, the GBR (2006) agreement has been widely applauded in the media and academic literature (Armstrong 2009;Dempsey 2011;Howlett et al. 2009). Several studies scrutinising the agreement have emphasised the innovativeness of remapping processes and outcomes. Thus McGee et al. (2010) explain the agreement's shift towards 'innovative collaborative planning', including the Joint Solutions Project between business and environmental non-government organisations (ENGOs); Affolderbach (2011) focuses on the role of ENGOs in brokering the agreement; Price et al. (2009) discuss the novel elements of the ecosystem-based management governing future logging; Low and Shaw (2011/12) highlight the new features of the agreement involving aboriginal peoples, notably government-to-government negotiations and the Coast Opportunity Funds; while Affolderbach et al. (2012) draw attention to the creation of a scientific boundary organisation, the Coast Information Team. The agreement featured fundamentally changed relations between environmental politics and science (Dempsey 2011), and is broadly influential, serving as model for the Boreal Forest Agreement of 2010 that covers 76 million hectares across Canada (Boychuk 2011).
This paper examines the process of innovation highlighted, if not problematised, in the studies cited above. Our study begins from Weiss's (2011) observation that institutional innovations are especially important in forestry. Weiss (2011, 11) broadly classifies institutional innovations as new cooperative organisations, laws, policies and procedures. In forest conflicts, such process innovations are designed to facilitate dialogue among stakeholders. If successful, they create and maintain durable 'products', organisations and routines that implement the goals of remapping and embody the compromises among stakeholders. These institutions vary in their functions, mandates, degrees of formality and permanence, and the scales at which they operate.
The GBR agreement comprises a set of related institutional innovations that collectively represent institutional thickening. Institutional thickness is a somewhat ambiguous and elusive concept, scarcely refined since North (1990) and Amin and Thrift (1994) identified institutional thickness as a key contributor to virtuous cycles of innovation observed in some specialised industries. The term evokes the new forms of governance and power topographies that underlie contemporary forestry, but it is not inevitable that remaps generate desirable local development. Rather, institutional innovation, thickening and remapping entail uncertainties that are both political and epistemological, rooted in the co-evolution of bargaining and learning among new and old stakeholders. Such processes are hard to predict, and institutional innovation is inevitably experimental, especially in peripheral regions re-inventing themselves. The engagement of more stakeholders in turn implies more diverse views and hopes for the future, with the future itself an ambiguous planning period. For the GBR, institutional thickening in support of forestry remapping is highly politicised, contesting definitions of resource values and development, while invoking experimental forms of governance that raise expectations about stakeholder behaviour and trust.
The specific goal of this paper is to better understand how the remapping of the GBR is working out in practice since the 2006 agreement. That agreement proposed a transition period, involving several land use orders for its implementation, and culminating in final legislation in February 2016 that now governs forest and environmental management in the region. More generally, this paper highlights the relationships between remapping, institutional innovation and thickening, not addressed by Weiss et al. (2011)  Institutional thickening and innovation 3 components of broader paradigmatic, Schumpeterian creative-destructive transformations in capitalist economies associated with the green techno-economic paradigm, sustainability transitions and related models (Coenen et al. 2012). Empirically, the second part of the paper draws on wide-ranging information sources, including interviews with stakeholder representatives, civil servants and industry consultants during 2015. It categorises stakeholder interactions on a bilateral basis as institutional innovations and assesses their implementation in the 2006-2015 transition period.

Stakeholders, remapping and innovation
Stakeholders are interest groups or institutions that influence or are affected by decision situations. In any particular situation, stakeholder interests in decision outcomes may not be matched by their influence. Indeed, a basic motivation for remapping forests is to empower stakeholders formerly marginalised in decisionmaking. Such empowerment, which simultaneously imposes constraints on vested interests, can itself be seen as an institutional innovation that democratises decision situations. In general, stakeholder models are analytically useful by revealing the basis for defining stakeholders and their relative powers, interests and motivations, and how stakeholders interact in conflict, cooperation and governance. The identification of stakeholders is not straightforward: in practical situations the inclusion of too many interested partiesa situation that arose in the first phase of the GBR negotiations (McGee et al. 2010, 751)can render decisionmaking ineffective. Parsimony is required in stakeholder designation, practically as well as conceptually, and stakeholder models necessarily will vary across the diversity of resource peripheries and conflicts. A four-legged-plus stakeholder model, based on government, business, community and ENGO designations for the BC context, provides the conceptual framework for this analysis ( Figure 2). The 'plus' refers to the courts as an entity within but distinct from government, and to First Nations as distinct communities within forest peripheries. This model also recognises that stakeholders at multiple scales are connected by learning and bargaining processes. These stakeholder categories are not homogenous or monolithic, and alternative categories are possible. Labour as an institution is noted but not highlighted, reflecting its diminishing power to shape remapping, and placed in the business category because of overlapping economic goals.
In BC, prior to 1980 and the onset of contemporary remapping processes, the dominant stakeholder interests were the provincial government and big business, with support from union labour. First Nations, ENGOs and the courts were not influential, and forestry science was dedicated to the industrial values of wood (Rajala 1998). With remapping, however, this simple, hierarchical model has become more messy and democratic.
First, national and regional governments (in federal jurisdictions) have pre-eminent roles in establishing the rules of the gamethat is, the institutions, policies and regulations that shape resource valuation and exploitation (Howlett et al. 2009). In Canada, provincial governments have enjoyed pre-eminent responsibilities for resource policies and have generally retained resources in public ownership while licensing private sector development rights. Indeed, around the world, forests were valued as industrial inputs with governments favouring economic development based on large-scale export-driven investments, stimulated by guaranteeing forestry corporations access to timber supplies, providing roads, railways, energy and even new towns, encouraging forestry science to meet industrial needs, and through the establishment of favourable regulations (Westoby 1989). BC's forest policy provides a paradigmatic example (Prudham 2007).
However, in recent decades governments have been challenged to give more priority to the non-industrial benefits of forests (Westoby 1989), notably in relation to conservation values and aboriginal rights. Indeed, the courts, with the power to compel government action, have been vital to the empowerment of Aboriginal Peoples in Australia (Moorcroft and Adams 2014) and Canada (Low and Shaw 2011/12) in ways that redefine resource property rights. In a series of decisions since 1997, however, the Supreme Court of Canada has 'revolutionized the jurisprudence of aboriginal rights and title' (Flanagan 2015, iii)itself labelled a 'remapping' by one legal authority (2015, 1)by rejecting the claim that aboriginal title had been extinguished, and by establishing the government's 'duty to consult' wherever claims of aboriginal rights or title are pending. In other provinces, treaties governing aboriginal rights and title were signed with many indigenous groups, but in BC few treaties were signed and more than 200 First Nations assert claims of aboriginal title, often overlapping, and covering most of the province. These court decisions established the basis for much stronger involvement by indigenous peoples in resource development (Borrows 2002), while potentially raising transaction costs for business decisionmaking.
Second, resource firms, especially large vertically and horizontally integrated corporations, have traditionally been the most significant vested interest exploiting forest peripheries. Even where pioneering small-scale enterprises have been historically important, the forest sector has witnessed significant integration tendencies. Moreover, the strategies of forest product corporations have been conservative and path-dependent, constrained by the need to modernise capital-intensive stages of production to remain competitive with rival oligopolists and by the specialised nature of managerial expertise and workforces (Edenhoffer and Hayter 2013a). For forest product corporations committed to commodity production, remapping has profound implications regarding resource access. In peripheries in the late stages of the resource cycle, these implications add to the uncertainties of corporate restructuring driven by technological change, increasing costs, declining competitiveness, recessionary conditions and even climate change. In BC's case, job losses, plant closures and divestments by foreign corporations in commodity production have been partly offset by an expanding population of smaller firms involved in secondary ('value-added') processing, along with a Canadianisation of the sector (Edenhoffer and Hayter 2013b). The value-added firms, however, have mainly located in metropolitan BC. Elsewhere, if surviving commodity mills remain important to local employment, remote forest communities have experienced decline.
Third, small, specialised forest resource communities typically evolved in close tandem with the needs and fortune of industry and dominant employers, ties especially evident in company (and union) towns that proliferate in many resource peripheries, such as BC or western Australia (Argent 2013;Markey et al. 2012). In more prosperous times, resource communities and dominant resource firms expressed little interest in diversification, but such attitudes are threatened by corporate restructuring and remapping. Indeed, when resource industries have downsized, especially if new residents immigrate for economic or lifestyle reasons, communities have become both more independent and interested in diversification (Brown 1995). In general, resource restructuring at least softens, if not fractures, community and worker support for vested industrial interests, often in association with more sympathy for environmental sustainability.
Moreover, aboriginal empowerment is in turn challenging traditional interpretations of the roles and composition of resource communities. In the Canadian literature on resource towns, stimulated by Lucas ENGOs comprise our fourth stakeholder group. Stimulating and responding to broad public concerns, the environmental movement has provided powerful advocacy of the non-industrial values of forests (and other resources) since the 1960s and especially since 1980. Significantly, the environmental movement has been important to reinforcing a strong social sense that

Figure 2 A four-legged-plus stakeholder model
Institutional thickening and innovation 'nature' is a public inheritance that provides multiple benefits and associated public responsibilities to ensure their sustainability. In the USA, for example, Joseph L. Sax led arguments to vindicate the legal claim that vital natural resources such as oceans, shorelines, air and some land, were a 'public trust' and needed to be conserved. Since 1970 this public trust doctrine, first adopted in Michigan, has spread throughout the USA, and many countries beyond, and helped fuel a rapidly growing environmental movement opposed to largescale resource commodification. For resources that remain under public control, the right to protest is particularly profound, and readily legitimised. ENGOs have been a powerful presence in forest conflicts around the world as they seek to modify or stop environmentally damaging logging and preserve the ecological values of old growth forests. Some ENGOs, such as Greenpeace, which originated in BC, are international in scope. However, ENGO activities are not about 'asset building' or 'market share', but through advocacy and communication seek to influence public opinion and ultimately public resource policies. ENGOs are media savvy and employ tactics ranging from political and consumer lobbying, logging blockades, shaming perceived environmental culprits, participation in committees at every scale of decisionmaking, coalition building and promoting ecological science. ENGOs vary in motives and tactics, but collectively they push governments to introduce policies that sustain environmental values. In BC, ENGOs have been a leading force in protesting forest commodity production, especially in relation to old growth temperate forests in coastal regions. After a decade of protest, they helped engineer an important agreement in 1993 promoting conservation of forests in Clayoquot Sound on Vancouver Island, which subsequently became a UNESCO biosphere reserve and world heritage site in 2000. ENGOs then shifted their attention north to the Great Bear Rainforest, an evocative and provocative name suggested by environmentalists (McAllister et al. 1997).
The shift towards a more democratic stakeholder model with new actors in forest conflicts implies more complex local-global dynamics. Sometimes in tandem and sometimes not, both ENGOs and indigenous peoples have sought to build alliances locally and globally, as they seek to influence public opinion, consumers and governments within BC and elsewhere. In Canada both ENGO and aboriginal opposition to vested forest interests have been reinforced by the ongoing trade dispute between Canada and the USA over lumber exports (Edenhoffer and Hayter 2013a). Thus US interests represented by the Coalition of Fair Canadian Lumber Imports (CFCLI) have argued that Canadian (especially BC) lumber exports are unfairly subsidised and have sought to restrict those exports since the 1980s. The CFCLI has become a watchdog or shadow stakeholder over BC forest policy, vigilant to any perceived Canadian violation. Over 30 years, the CFCLI, ENGOs and First Nations have often found common cause. Forest resource management, once considered a matter of provincial autonomy, has increasingly come under global scrutiny.
Forest conflicts are deep-seated and difficult to resolve for various reasons. The conflicts are normative, reflecting different beliefs about natural resource values. While industrial uses have dominated forest management, the ecosystem services, biocentric values and cultural significance of the forest are given priority by ENGOs and aboriginal peoples. Further, many nonwood benefits are intangible or non-economic and cannot be incorporated in quantitative trade-offs that underlie cost-benefit or multiple-use approaches. Remapping proponents have become skilled in opposing vested industrial interests, and organising effective, powerful alliances. Moreover, if remapping proponents have shared interests in opposing industry, they do not necessarily share the same aims, priorities or solutions (Hayter 2003).

Forestry remapping and paradigm change
Forestry remappings have occurred before as part of broader 'paradigmatic' restructurings of economy and society driven by technological and institutional innovations (Freeman and Louc ßa 2001). The link to technoeconomic paradigms helps illuminate the path-dependent challenge of contemporary remapping, the role of innovation and the stimulus of crisis. Drawing from western settler and especially North American experience, three broad forestry remappings can be identified in relation to broader paradigm shifts (Franklin et al. 1997;Hayes and Glendenning 2005). First, an era of deforestation  was part of large-scale 19thcentury industrialisation. A second remapping promoted a regulated 'scientific' forestry , a quintessential example of the Fordist paradigm. Since 1970, in a period variously referred to as post-Fordism, the era of flexibility or globalisation, forestry is being transformed by the principles of ecosystem management, with the hope that forestry will become part of a sustainable transition or green paradigm. These three periods of remapping respectively involved: colonisation, dispossession and deforestation; sustained yield and silvicultural forest management over large estates, privately owned or leased from the state; and adaptive forest management with multiple goals, with stronger commitments to conservation and community ownership. In broad terms, modern forestry evolved as a handmaiden to industry that became formalised as topdown, technocratically regulated 'scientific' practices that are now being challenged by concerns for more diverse values. If this scenario is found in North American and western settler societies, it resonates elsewhere, such as in the Himalaya (Guha 2000).
A key part of the present paradigm shift, deeply implicated in forest conflicts, is the debate over scientific wisdom in forestry practices (Langston 2005;Rajala 1998). Thus modern forestry science and practices were developed in Europe, mainly Germany, and became similarly institutionalised in the North American forestry profession whose 'rational' advice 'imposed a rational, uniform and simplistic order on the complexities of localized ecological systems' (Lee and Field 2005, 3). For Langston, this advice became an intellectually authoritarian framework that defined sustained yield and silviculture according to industry's needs. This thinking, for example, equated old growth forests, that is forests in which productivity gains and losses offset each other, with decadent timber, increasingly worthless to industry. Indeed, as recently as 1986 a leading forest economist labelled BC's remaining old growth forests in exactly this way (Percy 1986). In this view, old growth forests should be logged as fast as possible and replaced by newer, more productive forests, commonly in single-species single-aged stands. Community benefits and healthy forests were consequences assumed to flow from this rational management system.
However, as Langston (2005) shows, conventional forest science orthodoxy often failed in its own terms. Changing the species mix led to more damage by disease and insects, and increased vulnerability to fire; even-aged clear-cutting led to similar problems and conflicted with other users; and mills were promised optimistic timber allocations, making mill closures inevitable. Meanwhile the ecological values of oldgrowth forests were ignored. For proponents of remapping, however, these values are central, and were underlined by alternative research and forms of knowledge not mandated to grow trees solely for wood production. For Langston, forest conflicts in Oregon stimulated the various stakeholders to engage in understanding each other's perspectives: What mattered most about litigation was that it forces a variety of stakeholders, with multiple voices, multiple stories, and multiple perspectives to communicate with each other. (2005,72) Co-evolution of bargaining and learning is not straightforward. Thus the critics of conventional forestry practices not only point to alternative scientific evidence but emphasise local variations in ecological processes and that the scientific understanding of dynamic ecosystems is constantly changing, never complete. Drawing on Firey's (2005Firey's ( [1963) insights, this debate is rooted in future expectations that are judgemental, varied and constrained: remapping requires agreement over the stable institutionalisation of future-referring values that are 'space and timebound: they are tied to particular social orders whose eventual demise they are destined to share ' (2005 [1963], 27). Within these bounds, remapping depends on stakeholders establishing mutual trust over recognising multiple values, a shared willingness to sacrifice some present for future values, and shared sense of obligation to local territory. There are no mechanical formulas to achieve such cooperative behaviour.
Stakeholder-driven remapping stresses the need to make policy choices based on a diversity of scientific arguments and knowledge; in Weiss' (2011) terms, innovations that change the political-institutional framework of forestry are required. Innovation is conventionally divided in a business context into product, process, marketing and organisational innovations, varying from incremental to radical. However, Weiss argues that institutional innovations that are hard or impossible for market actors to initiate are important in forestry and relate to changes in procedures, regulations or incentives or joint actions by public or semi-public organisations. Further, in stakeholder models, innovation implies inter-organisational cooperation that may be horizontal or vertical and may be seen both as an innovation itself (a new organizational model) and as a supporting factor for innovations (e.g. information exchange in sector meetings or maintaining cross-sectoral relations). (Weiss 2011, 15) Such supporting process innovations are particularly significant in forest conflict situations as a condition for achieving cooperation over integrating disparate knowledge bases and epistemologies among stakeholders prior to negotiated agreement(s). As we note below, a scientific boundary organisation illustrates such an innovation that has been key to the GBR negotiations. The hope is to create durable institutions to organise resource governance. Admittedly, the distinction between facilitative (supporting or process) and outcome (product) institutional innovations can be blurred, especially in light of their experimental nature. However, this messiness underlines the need to analyse related innovations that comprise institutional thickening.

Stakeholder collaboration and institutional innovation in the Great Bear Rainforest
The second half of this paper explores to what extent the institutional thickening produced by the GBR remapping can generate a virtuous circle of innovation. The four-legged-plus stakeholder model generates six possible bilateral avenues for conflict resolution and the development of cooperation (Figure 3). This schema represents the initiatives, practices and organisations that have unfolded since 2006. While these interactions are represented as bilateral because they were generated initially to solve bilateral conflicts, they occur within a multi-stakeholder dynamic and an environmental governance model in which collaboration, or at least acquiescence, of all parties is necessary to validate and implement agreements. These new institutions include a firm, a foundation, an alliance, a protocol and a set of management practices. They also include changes to existing institutions, such as those regulating forest tenure and logging practices, significant financial investment in conservation and development projects and aboriginal capacities to monitor, conserve and develop resources in their traditional territories. Collectively they reflect significant institutional thickening.
The empirical analysis draws on research conducted over the past 15 years on the GBR, including interviews, participant observation, literature review and geospatial analysis. Earlier research focused on understanding the origins of the conflict, the nature of bargaining and hopes for remapping. This paper shifts attention to outcomes, and derives from new research, featuring ten interviews conducted in the summer of 2015 with ENGO representatives, civil servants, forest industry consultants and aboriginal representatives with long experience in GBR negotiations, and who remain directly involved in outcomes. Our questions were open-ended but focused on how aspects of GBR plans and agreement are working out in practice. Indeed, the collective understanding and interactions of our 2015 respondents are a significant part of the social capital and institutional thickness on which regional development will depend going forward. As prelude to the six bilateral relationships, the key facilitative or supportive innovation is briefly referenced.

Facilitating the GBR agreement: the Coast Information Team
A scientific boundary organisation, the Coast Information Team (CIT), was a key facilitative development for the GBR agreement, established to separate disputes over science from interest-based negotiations (Clapp and Mortenson 2011). The CIT's mandate was to establish a common knowledge base on which all stakeholders agreed, to answer questions posed by negotiators and to support the bargaining process (Affolderbach et al. 2012). The shared knowledge base generated through this process was a GIS of the region that mapped multiple cultural, economic and ecological characteristics. The CIT itself emerged as a compromise in 2001 after lengthy and often heated interactions among the stakeholders during the regional planning process initiated by the provincial government in the 1990s. That is, institutional thickening occurred as a precondition to agreement.
In turn the CIT provided the foundation for the GBR agreement of 2006 that identified extensive new protected areas, established land use zoning at multiple scales and committed to ecosystem-based management throughout the forest matrix. The CIT has since been disbanded, but its work outputs, notably the ecological spatial analysis, have been essential for the development of ecosystem-based management discussed below. The same negotiations also generated more permanent institutions, notably an agreement establishing the Coast Opportunity Funds with public and philanthropic capital to support conservation-based development. In general terms, this agreement illustrates the paradigmatic changes from government to governance as they occur in forestry decision-making, and its shift from a top-down model to a more democratic model in which state and big business powers are diluted and new social forces are recognised as stakeholders. This example illustrates the role of conflict in stimulating the innovations underlying paradigm change. We argue that without the intense forestry conflicts that exploded in BC in the 1980s and 1990s, such a remapping agreement could not have happened. Vested interests would have had no incentive to change behaviour, and public consciousness located in distant metropoles would scarcely have noticed. Such a view echoes the claim by Lee (1984) that sustained yield principles were developed in Germany in the 19th century, not out of extant cooperative behaviour as commonly supposed, but to resolve highly problematic, unstable situations.
However, the relationships between institutional innovation, institutional thickness and conflict are problematical. Conflict and protest may stiffen the opposition from vested interests, place-based targeting by interest groups may lead to unintended consequences elsewhere, antagonism may delay compromise, business may prefer to invest elsewhere, and the energies used in protest and conflict resolution may be exhausted, no longer available for constructive engagement. Moreover, remapping agreements are plans that need to be implemented; plans may or may not be executed. Indeed, a key question for remapping agreements that are derived from high levels of mistrust and conflict is whether they have actually become institutionalised, that is become the basis for durable, stable behaviour. The construction and maintenance of social capital is not a guaranteed consequence of institutional thickening.

Business and ENGOs: the Joint Solutions Project
Collaboration between industrial and environmental interests has erected a new framework of institutions to implement 'socially responsible' ideas of self-regulation without direct government intervention. Among these institutions are at least three types of firms: valueadded processors, third-party verifiers and multinational retailers who market and demand certified wood products (e.g. IKEA, Home Depot and others). More important for the Great Bear Rainforest agreement, however, was a bilateral collaboration that emerged during multilateral negotiations in the summer of 2000.
To break the impasse created by continued logging and the market campaigns, four forest companies (Interfor, NorskeCanada, Western Forest Products and Weyerhaeuser) began negotiations with four ENGOs (Forest Ethics, Greenpeace, Rainforest Action Network and the Sierra Club), leading to a suspension of market campaigns against the companies in return for a moratorium on logging in 30 watersheds critical for conservation. The resulting alliance was dubbed the Joint Solutions Project (JSP), and its proposals broke a logjam in negotiations in 2001(Mortenson 2005. The bilateral agreement restarted the multi-sector negotiations that would result in an interim land use map, the establishment of the Coast Information Team, and commitments to apply ecosystem-based management in the coastal forest matrix. The JSP negotiated internal agreements over logging practices, protection areas and development time frames based on compromises acceptable to both parties, and eventually other stakeholders as well. The JSP continued to play an essential role beyond the initial compromise. The best evidence of its influence has been to reach agreement that under ecosystem-based management, protection would be set at a precautionary 70 per cent of the range of natural variation within the operating land base. This Interviews with ENGO and industry representatives confirmed that this bilateral collaboration continued to facilitate communication and to negotiate compromises reflected in memoranda of understanding. The JSP by itself, however, does not demonstrate institutional thickening because it was created for negotiations that have been concluded with the formal 2015 agreement. In the words of an industry consultant, the 'JSP has run its course' (Interview 1). Institutional thickness that promotes regional development is described in the literature as existing in ongoing practices. The JSP helped establish the groundwork for broader collaboration, but like the Coast Information Team, may itself cease to exist.

British Columbia and First Nations: government-togovernment negotiations
Cooperation and convergence between business and ENGOs reflect voluntary commitment and self-regulation within civil society, a phenomenon also evinced by eco-certification bodies like the Forest Stewardship Council. Other aspects of institutional thickening are directly related to process innovation in formal government, contrasted with broader and more diffuse processes of governance. Political and cultural interests have reserved final assent to stakeholder-negotiated agreements to the province and First Nations in a subsequent and higher level of government-to-government (G2G) negotiations.
G2G negotiations emerged from the initial 2001 stakeholder consensus recommendations to government. The notion of a subsequent and higher level of review was important to First Nations because it recognised at least the possibility of aboriginal title and shared sovereignty over natural resources. G2G negotiation was predicated on an agreement, the Turning Point Initiative, that brought the 27 different aboriginal groups with traditional territories in the region to develop common conservation, cultural and economic goals and to participate in both provincial and federal negotiations with one voice as Coastal First Nations (coastalfirstnations.ca; Davis 2009).
First Nations' new influence, including veto power over stakeholder recommendations, arose from a series of decisions by the Supreme Court of Canada establishing the existence, parameters and obligations of aboriginal title, beginning in 1997 with Delgamuukw v. British Columbia. The decision confirmed that aboriginal title had not been extinguished in BC, and that the Crown was obliged to consult and accommodate First Nations. While G2G had been mooted in the early 1990s, the Delgamuukw decision and Great Bear Rainforest negotiations marked the shift of G2G negotiations from an ill-defined concept to a functioning institutional structure, as well as the 'shift in the role of First Nations: from a particular kind of "stakeholder" to a long-term government partner' (Barry 2012, 217).
G2G negotiations represent both an advance and a complication of aboriginal relations within BC. It is a victory for First Nations because it formally recognises their right to be consulted not merely as stakeholders but potentially as a quasi-sovereign government with unextinguished aboriginal title. For the province, it constitutes a rapprochement in a long and largely fruitless process of treaty negotiations. First Nations now exert control over where and how much logging can take place, but the limitations of G2G negotiations are keenly felt: the protocol allows for First Nations review of and consent to logging operations, but not pipelines, fish farms or grizzly bear hunting (Interview 2; Interview 8). Several respondents independently characterised the province's approach to G2G as one of 'containment', resisting attempts to extend the model elsewhere in the province.

Government and ENGOs: ecosystem-based management
Government and ENGOs have found productive, if conflictual, engagement in defining ecosystem-based management (EBM) for the GBR. EBM emerged as a vital component of the compromises that began in the Joint Solutions Project. EBM was essential to the 2001 interim agreement because it assured ENGOs that ecosystem function would be better maintained than in conventional forestry by extending ecological priorities beyond protected areas to the entire forest matrix. EBM was approved by the province and First Nations in G2G negotiations, and now guide the management of the forest matrix outside the new protected areas. Nevertheless, reaching agreement on the definition and implementation of EBM has taken more than a decade of painstaking analysis and negotiation (Interview 4).
EBM sets goals and objectives at multiple scales: at the sub-region or territory (500 000-5 000 000 hectares), the landscape (30 000-100 000 ha), the watershed (1000-50 000 ha) and the site (under 250 ha). To measure these objectives for different ecosystem types under varying scenarios of logging pattern and intensity, EBM is constructed using an Ecosystem Spatial Analysis generated to inform the stakeholders at the negotiation table. The spatial analysis is a sharedsource database in which all stakeholders' technical specialists agreed to the coordinate system. It provided a detailed mapping of ecosystems and enabled their characterisation by regional rarity as well as their economic accessibility and timber volumes (Clapp and Mortenson 2011).
Setting protection targets depends on a body of science sufficient to determine for each constituent ecosystem within the region a 'range of natural variability', defined as the range of dynamic change in natural systems over historic time periods (Allen 2005). The Ecological Spatial Analysis was essential to the GBR agreement: it allowed delicate compromises to be worked out with confidence that they could ultimately be implemented. For example, in the 2003 agreement, the stakeholders agreed on a target for old-growth forest retention at the landscape level of 50 percent of the natural proportion, provided the average across all landscapes is 70 percent (the sub-regional target); and at the watershed level 30 percent, provided the average across all watersheds is 50 percent (the landscape level target). (Allen 2005, 11) The final protection level proposed by the JSP in 2014 and legislated by the province in February 2016 was a more straightforward 70 per cent of the range of natural variability. The JSP commissioned a literature review and meta-analysis of 20 cases of ecosystem resilience or collapse (Price et al. 2007). Ecological risk was theorised as the area of habitat loss relative to a range of natural variation. More than 60 per cent remaining intact of the natural benchmark projected low risk to biodiversity, while below 30 per cent of the benchmark risked the onset of rapid decline with effects cascading through multiple species and the food chain generally (Interview 2). The science was enough for a negotiated outcome: the JSP ultimately agreed that ecosystem-based management should mean protecting a precautionary 70 per cent of the matrix, which one industry consultant characterised as leaving only 15 per cent of the productive area available for harvesting (Interview 1).
Although EBM is an ongoing government function, its key parameters have largely been set by business and ENGOs. The province is wary of establishing EBM as a precedent for other regions, regarding it as researchintensive (Interview 2), expensive (Interview 3) and impinging on provincial sovereignty (Interview 4). Several respondents characterised government policy on both EBM and G2G negotiations as 'containment'concessions necessary to achieve peace in the Great Bear Rainforest, but not models for the rest of the province. Stakeholder processes begin with government delegating some of its powers, and First Nations and ENGOs have gained substantial influence at the cost of governmental discretion and autonomy. The province seems determined to avoid extending a GBR model of G2G negotiations and ecosystem-based management.

First Nations and ENGOs: the Coast Opportunity Funds
Community and environmental collaborations are reflected in informal, occasionally uneasy, alliances during the land-use zoning negotiations. Informal alliances may not lead to permanent institutions but nevertheless affect negotiations at critical junctures, and are key elements of social capital. Whether those informal collaborations continue in the absence of a formal institution charged with promoting them is an important question, but empirically difficult to answer. At least one aspect of those collaborations has been formalised in the Coast Opportunity Funds.
One of the key compromises in the GBR agreement was the contribution of capital to fund conservationbased development proposals. A frequent challenge from communities and First Nations to ENGOs during the negotiations was to 'put their money where their mouths are', calling on environmentalists to find resources to make up for the logging jobs that would be lost in any reduction to sustainable levels (Mortenson 2005). The Coast Opportunity Funds were established in partnership between BC, Canada and private donors, notably the international foundations that were instrumental in funding the ENGO market campaigns, among them the Rockefeller Brothers Fund, the Hewlett Foundation, the Packard Foundation and the Pew Charitable Trusts (Tedesco 2015). Of the Funds' CAD 120 million in initial funding, half came from the foundations, and the other half from the provincial and federal governments (Davis 2009). The Great Bear Rainforest is widely agreed to be of global conservation significance; the Coast Opportunity Funds muster global resources to fund its conservation. An industry observer commented that it was nice to see the foundations supporting local capacity instead of market campaigns (Interview 1).
The Coast Opportunity Funds comprise two separate funds: the Conservation Fund supports participating First Nations by funding projects to ensure the ecological integrity of the GBR, such as ecological research; habitat protection, restoration and enhancement; and capacity-building for resource planning, management and monitoring. The Conservation Fund is a permanent endowment of CAD 56 million, with annual expenditures of approximately CAD 2 million (Coast Opportunity Funds 2015). Perhaps the most significant achievements of the Conservation Fund have been the establishment of integrated stewardship offices for many participating First Nations (Interview 7), enabling them to decide which and what kind of development proposals they wish to allow. Prior to 2007, some First Nations had fisheries departments. Five years later most had broadened their focus to include forestry and ecotourism proposals consistent with the conservation aims of the GBR agreement. Some First Nations resource management offices now review companies' logging plans before they are submitted to the province, achieving the objectives of both business and community more efficiently.
Equally significant are the newly created Guardian Watchmen programmes. Based on the model of the Haida Watchmen Program, formed in 1993 to monitor fisheries and cultural heritage sites in the islands of Haida Gwaii, Coastal First Nations have promoted a network of similar organisations for the central and north coasts. Monitoring is an ongoing cost of environmental planning, one that is often de-funded when government cuts costs. Support from the Coast Opportunity Funds has enabled First Nations to establish their own monitoring programmes, and helped to leverage additional funding from other philanthropic foundations (Interview 7).
In contrast to the permanent endowment of the Conservation Fund, the Economic Development Fund of CAD 58 million is to be spent in its entirety by 2017 to support sustainable businesses and community-based jobs in the region (Coast Opportunity Funds 2015). The economic development funds are allocated (in confidential discussions) to First Nations in varying proportions, with larger shares of the funding allocated to First Nations whose traditional territories contained larger shares of the newly established conservancies (Interview 7). The Coast Opportunity Funds have been instrumental in establishing new institutions and capacity within the First Nations, building up institutional thickness at the regional and territorial level, with ongoing support that promises to survive beyond the conclusion of the negotiations.

First Nations and Business: Coast Tsimshian Resources
Some further evidence of institutional thickening is seen in the growing connections between industrial and aboriginal interests. At least ten First Nations in the GBR have signed revenue-sharing agreements with forest companies operating in their traditional territories (Hoekstra and Pynn 2015), although the amount and proportion of revenue shared remains confidential. Perhaps more significant is the emergence of First Nations-owned companies. The most prominent example is Coast Tsimshian Resources, which acquired the forest tenures of the bankrupt Skeena Cellulose in 2007, with long-term annual harvesting rights of 564 314 cubic metres (Hoekstra and Pynn 2015). Coast Tsimshian Resources employs 110 workers, maintains a marketing office in Shanghai and has pioneered the export of hemlock logs from the north and central coasts to China, Japan and Korea.
Aboriginal ownership or consent is requisite to ecocertification by the Forest Stewardship Council, which includes assurance of indigenous peoples' consultation, accommodation and participation. This innovation is important for expanding use of coastal western hemlock (Tsuga heterophylla, also known as green hemlock), a lower-quality but still abundant resource that has lacked markets, evincing the expanding participation of First Nations in the forest industry.

Government and Business: forest tenure, community forests and conservancies
Reform of the longstanding political-industrial alliance between big business and the provincial government can be seen in the retrenchment of forest tenure arrangements, including the expansion of log auctions, the clawback of forest tenures and the creation of new community forests. These are not necessarily institutional innovations: they may better be described as new priorities expressed through existing institutions, such as when the Chief Forester reduced the Annual Allowable Cut for the Mid Coast Timber Supply Area to 767 000 cubic metres in 2011 to reflect the withdrawal of new conservancies from the working forest (Ministry of Forests, Lands and Natural Resource Operations 2011). Nevertheless, the GBR has seen the expansion of the community forest model A final institutional innovation reflecting ongoing collaboration between government and business can be seen in the conservancies, a new category of protected areas created in the Great Bear Rainforest (Stronghill et al. 2015). Conservancies limit logging, hydro-electric development and other large-scale industrial developments, but allow exceptions for aboriginal commercial activities, as well as subsistence and ceremonial activities. While logging is restricted, fishing, aquaculture, guiding, filming, log salvage and the harvest of nontimber forest products are included in various conservancy management plans (Stronghill et al. 2015). To keep a campaign promise made to the mining industry, the provincial government insisted that new protected areas should not place geologically prospective areas off limits to mineral exploration and development. The conservancies are to be co-managed by First Nations and the province.

Conclusion: institutional innovation and paradigm change
The GBR has moved from government to governance, in which the role of the state is diffused and dispersed among multiple stakeholders. The most significant change in the new power topology is the provincial state, which has obtained peace in the woods at the cost of significant autonomy over resource allocation, conservation and management. Stakeholder negotiation tables generated the agreements for the province's ratification. Under the terms of G2G negotiations, ratification was shared with First Nations.
Of the six bilateral interactions in the four-leggedplus stakeholder model, most resulted in rapprochement or compromise, the removal of obstacles to industrial operations and the expansion of community control. Each had elements of novelty and local adaptation, while drawing on precedents and models from other resource peripheries. If aboriginally owned firms, bilateral alliances, community forests, conservation foundations, ecosystem-based management and G2G negotiations have precedents elsewhere, they collectively represent institutional thickening in many dimensions on an unprecedented scale in this thinly populated region. Changes to forest tenure in the context of remapping environmental and cultural values also represent a significant shift of stakeholder power, supporting the characterisation of a new paradigm in forestry. The outcome of this remapping process is neither a free market solution, nor public control, but an increasingly complex architecture of institutions, based in both civil society and the state, that promote sustainability, resilience and legitimacy. This paper has argued that regional development is structured by institutions constructed and thickened during conflict resolution. New or modified institutions are needed to implement stakeholder compromises. Process innovation facilitates bargaining, drives learning and secures the social licence to operate. Despite the emergence of a new paradigm, the development of shared norms, values and behaviours is not an automatic consequence of new maps and formal agreements: the efficacy and the survival of each of these institutions remain uncertain and contingent on followthrough, goodwill and the transfer to productive engagement of the social network constructed during negotiations. Government-to-government negotiations are a process innovation in the BC context, but how or even whether they will continue in the GBR or elsewhere in the province remains unclear.
Furthermore, process innovation in remapping is as much associated with satisfying gatekeeping conditions and securing the social licence to produce, as it is with increasing productivity or profitability. Innovation in the context of a new paradigm in contested resource peripheries is far removed from the notion of science as handmaiden to industry. Many of the new collaborations and alliances have been generated by the need to seek public legitimacy because survival rather than increased efficiency was at stake. This poses a challenge to the teleological spirit of regional development theory: how can we tell if institutional thickening constitutes a positive development, especially where process innovation is concerned? Our paper has demonstrated substantial institutional thickening in the GBR, but not necessarily a virtuous circle of innovation along the lines theorised by North (1990), Amin and Thrift (1994) and others. Whether these new institutions reflect the formation of enduring social capital that promotes regional development, or a sclerotic hyper-regulation that hinders it, requires further research.