Social License to Operate: Legitimacy by Another Name?

Social license to operate is an increasingly used but seldom defined concept. In this article, we draw from academic, popular, and industry literature to identify and synthesize three models of social license to operate. Building on our review, we investigate the linkages between social license to operate and legitimacy and consider how the two concepts differ from and interrelate with one another. Then, we review the various methods that have been used to measure social license to operate. We conclude by discussing the implications for stakeholder engagement, evolving models of regulation, and potential avenues for future research.

concept, we aim to foster greater scholarly appreciation of the key concepts and diverse frameworks potentially implicated in discussions of social license to operate. Notably, our review identifies and synthesizes three different varieties of SLO. After highlighting some of the similarities and differences among these varieties, we investigate the linkages between SLO and legitimacy, paying attention to how the two concepts differ from and interrelate with one another. We then review methods that have been used to measure social license. Overall, our review demonstrates opportunities for more systematic and nuanced terminological use, while suggesting the need for further empirical and theoretical research. We close by discussing implications for stakeholder engagement, evolving models of regulation, and potential avenues for research.

Varieties of social license
The pyramid model The first variety of social license, the pyramid model, was developed iteratively in a series of articles, papers, and presentations by mining industry consultants over the 14-year period from 2000 to 2013. Proponents credit James Cooney, then an executive at Canadian gold mining company Placer Dome, for inspiring the model (for reviews, see Black 2013; Boutilier 2014; . 3 In 1996, Placer Dome had been severely criticized after a tailings dam failed at one of its mines in the Philippines, releasing toxic mud into a river and burying a village (Boutilier 2014). More generally, mining was ranked the worst of 24 U.S. industries in a 1996 Roper opinion poll, behind even the tobacco industry . It was in this context in 1997 that Cooney reportedly characterized the industry's problems to World Bank officials as a matter of obtaining a "social license to operate." World Bank personnel are said to have circulated the term at a May 1997 conference on mining and the community .
A February 2000 article in CIM Bulletin entitled "Earning a Social License to Operate" (Joyce and Thomson 2000) appears to be the first attempt to provide the term with some conceptual substance. Joyce, a consultant with Calgary, Canada-based Golder Associates, and Thomson, an independent consultant from Vancouver, Canada, began their article by cataloguing various "social risks" facing mining companies in Latin America. At the project level, these risks were argued to threaten social acceptability by posing what they termed "problems of legitimacy" (Joyce and Thomson 2000: 51). They explained: A social license to operate exists when a mineral exploration or mining project is seen as having the approval, the broad acceptance of society to conduct its activities. . .Such acceptability must be achieved on many levels, but it must begin with, and be firmly grounded in, the social acceptance of the resource development by local communities (Joyce and Thomson 2000: 52).
In a 2002 landmark industry report, the International Institute of Environment and Development declared that the industry had "failed to convince some of its constituents and stakeholders that it has the 'social license to operate' in many parts of the world" (International Institute of Environment and Development 2002: xiv as quoted in Owen and Kemp 2013: 29). Without defining what it meant by social license, the report revealed an inextricable link with the mining industry's survival. In short, this variant of social license to operate "emerged as an industry response to opposition and a mechanism to ensure the viability of the sector" (Owen and Kemp 2013: 29).
Usage in the industry became widespread thereafter. In a 2003 project sponsored by Newmont Mining Corporation, Business for Social Responsibility (BSR) published six case studies on social license in the mining industry, defining the concept as "gaining support for the project from concerned groups, or stakeholders, over and above meeting any legal requirements" (BSR 2003: 4). A convenience survey (n 5 152) of mining industry participants conducted in 2005-2006 revealed that 78 per cent of respondents were familiar with the term "social license." 4 Whether previously familiar with the term or not, 81 per cent agreed that, intuitively, social license was an apt way to describe societal and community acceptance of the right to access and extract mineral resources; and 56 per cent reported that their organizations used the term or its underlying concepts (Nelsen 2007: 9-10). Looking back on these and other early efforts, scholars have concluded that the concept of social license to operate initially emerged as little more than a memorable turn of phrase, "a term largely invented by business, for business" (Morrison 2014: 14; see also Raufflet et al. 2013).
Some years later, in an industry conference presentation, Thomson and Joyce (2008) expanded their definition of social license to operate to include three "normative components:" legitimacy, defined as "conforming to established norms. . .legal, social, cultural and both formal and informal;" credibility, defined as "the quality of being believed-the capacity or power to elicit belief;" and trust, defined as the "willingness to be vulnerable to risk or loss through the actions of another." 5 They further differentiated between project acceptance and approval, arguing that legitimacy is necessary for acceptance, but credibility and trust are necessary for approval. Boutilier (2014), also an industry consultant, credited Joyce and Thomson for: (a) being the first to define social license to operate in terms of legitimacy; and (b) proposing that social license to operate at the project level would promote reputation benefits at the corporate level. 6 see also Boutilier and Thomson 2011;Thomson, Boutilier and Black 2011) later elaborated their definition into a multilevel pyramid model (see Figure 1). In this model, legitimacy distinguishes projects that have been rejected (that is, projects for which social license to operate has been withheld/withdrawn) from those that have been accepted by stakeholders through engagement with them according to the "rules of the game." Credibility distinguishes projects that have been accepted from those that have been approved by stakeholders through formal negotiation, definition, and agreement on the roles and responsibilities of the company and stakeholders. Finally, trust distinguishes projects that have been approved from those for which stakeholders have adopted what they called a sense of co-ownership or psychological identification through collaborations, shared experiences, and vulnerabilities. More recently, the Australian Centre for Corporate Social Responsibility (ACCSR) adopted the pyramid model of social license to operate. Notably, the centre's managing director, Leeora Black (2013), has promoted the pyramid model as a "management framework for complex times." The three strand model A second variety of social license, what we refer to as the three strand model, was developed in a series of interrelated publications in 2003 and 2004 Thornton 2003, 2004;Kagan, Gunningham and Thornton 2003;Thornton, Kagan and Gunningham 2003). Whereas the originators of the pyramid model conceived of "social license to operate" as something of a dependent variable, which was in need of explanation, Gunningham and colleagues (Gunningham, Kagan and Thornton 2003: 2) started with a series of research questions: "Why has corporate environmental performance improved over time? Despite this improvement, why are some firms better environmental performers than others? How, and to what extent, can corporations be motivated to go beyond compliance with existing environmental regulations?" Ultimately, these research questions resulted in the inductive emergence of social license as one part of a larger explanatory framework. In other words, social license to operate emerged as something of an independent variable critical to explaining why some companies went beyond merely complying with environmental regulations, while other companies fell short of regulatory compliance.
To answer their research questions, these scholars conducted an indepth study of the environmental performance of 14 pulp mills located in the United States, Canada, Australia and New Zealand. Based on their findings, the authors proposed that companies in "closely watched industries" depend upon a multi-stranded license to operate: legal license, or the regulatory permits and statutory obligations embodying the demands of regulators, legislators, and judges; social license, or the demands of local, national, and international environmental activists, local community groups, and sometimes the general public; and economic license, or the profitability demands of top managers, lenders, and investors. In addition to their direct effects, the authors proposed that these different strands have interactive effects. Environmental groups may seek to enforce social license directly (for example, through shaming and adverse publicity), but also may attempt to influence economic license (for example, by generating consumer boycotts of environmentally damaging products) and legal license (for example, through citizen lawsuits or political pressure for regulatory initiatives).
More recently, John Morrison (2014), executive director of the Institute for Human Rights and Business, proposed a variant of the three strand model (see Figure 2) that substitutes political license (that is, the authority that the government gives to any other organization to undertake a particular activity) for economic license (see also Brueckner et al. 2014 on the close link between politics and economics in the context of SLO). Although this may seem like a significant difference, it may be of little consequence. According to Morrison, the only time political license is not driven by economic considerations is if "you live in North Korea" (Morrison 2014: 22).
Over the past decade, the original three strand model Thornton 2003, 2004;Kagan, Gunningham and Thornton 2003;Thornton, Kagan and Gunningham 2003) has spawned a variety of followon work. Lynch-Wood and Williamson (2007) examined whether social license concerns are sufficient to entice smaller firms to go beyond regulatory compliance in their environmental performance. They identified five factors that comprise social license-environmental impact of the firm's products and processes; customer power; customer interest; corporate/ brand visibility; and community pressure-and argued that at least two of these factors must be salient for a small and medium enterprise (SME) to go beyond compliance. They concluded that for most SMEs, none of these five factors are significant enough to encourage going beyond compliance, and therefore, regulators cannot depend on social license considerations to either incentivize or sanction these firms. Thornton, Gunningham, and Kagan corroborated this conclusion in two studies of the U.S. trucking sector Gunningham 2003, 2009). In these works, social license was defined as pressure from communities, advocacy groups, employees, and the news media. They concluded that, due primarily to low social visibility, environmental decisions in small trucking firms are driven almost entirely by economic license, and that social license pressures faced by these firms are very weak.
Howard-Grenville, Nash, and Coglianese (2008), contributed to this discussion by showing that organizational activities beyond compliance are not solely driven by external factors, such as social license concerns. They examined the impact of five internal factors (managerial incentives, organizational culture, organizational identity, organizational self-monitoring, and personal affiliations and commitments) in 10 companies, five of which were participating in the U.S. Environmental Protection Agency's voluntary National Environmental Performance Track program. The authors found that firms in the two groups differed on company identity, selfmonitoring, and managerial incentives and concluded that along with external regulatory, social and economic factors, internal factors significantly influence a company's willingness to go beyond compliance.

The triangle model
The third variant of social license, the triangle model, is based on the concept of social acceptance that emerged in the 1970s and 1980s amid the first concerted efforts to develop renewable energy policies. Whereas the pyramid model conceives of social license to operate as an outcome worthy of explanation and the three strand model conceives of social license as part of the explanation for environmental performance that goes beyond regulatory requirements, the triangle model conceives of social license as resulting from a threefold set of "acceptance processes." Namely, these researchers conceived of social acceptance as building confidence, familiarity, and trust in environmentally-friendly, but unproven technologies. Social acceptance is considered necessary to generate policy maker support for the financial and regulatory incentives required to overcome entrenched interests and the path dependency of conventional fossil fuel energy systems.
Initially, the problem of social acceptance was largely neglected, in part because public opinion surveys indicated very high levels of support for renewable energy options. However, Carlman (1984) showed that public opinion surveys did not necessarily translate into public, political, and regulatory acceptance of renewables, such as wind power. More recently, W€ ustenhagen, Wolsink andB€ urer (2007: 2683) highlighted social acceptance as "a powerful barrier to the achievement of renewable energy targets." As shown in Figure 3, they distinguished three dimensions: sociopolitical acceptance, or the broadest, most general level of social acceptance of both policies and technologies by the public, key stakeholders (that is, employees) and policymakers (see also Jegen and Philion 2017); community acceptance, or "the specific acceptance of siting decisions and renewable energy projects by local stakeholders, particularly local residents and local authorities" (W€ ustenhagen, Wolsink andB€ urer 2007: 2685), which may vary over time; and market acceptance, or the process of widespread adoption of an innovation. In particular, energy projects are embedded in complex multisided infrastructures that involve consumers, investors, and producers.
This three-pronged distinction continues to provide a useful point of departure for this literature. Public opinion polls tend to show that while the general public remains favourable to the idea of wind, solar, biomass, wave, geothermal, and other renewable energy technologies (socio-political acceptance), host communities are not as supportive (community acceptance) (Devine-Wright 2011; Pasqualetti 2011). Findings show evidence of a "social gap" (Bell et al. 2013;Bell, Gray and Haggett 2005) between public support for the general goal of more wind energy and the level of local support for specific projects.
Policymakers, renewable energy developers, and other experts tend to view those who oppose local renewable energy projects and adopt a not-inmy-backyard (NIMBY) position as ignorant and uninformed. Evidence indeed shows that some members of host communities take the NIMBY position (Walker, Baxter and Oulette 2014;Wolsink 2000Wolsink , 2007. However, researchers strive to understand opponents on their own terms (Aitken 2010;Devine-Wright 2011), and some (for example, Cohen, Reichl and Schmidthaler 2014) have argued that NIMBY is a rational reaction that should be recognized and remedied through policies that compensate individuals for costs associated with living close to renewable energy projects. Batel et al. (2013) made important conceptual critiques of the term "social acceptance," pointing out that scholars tend to erroneously conflate acceptance with support. They showed that residents react differently when asked if they would accept vs. support an energy project. "Acceptance" implies a more passive relationship to an energy project. Likewise, Walker and Cass (2007) observed that scholars (and policymakers) tend to view members of the public as falling on a spectrum between support for and opposition to such projects; however, members of the public also play other roles, as service users, investors, owners, lessors/ lessees of land or buildings, captive consumers, protestors and more. The point is that acceptance and support implies a degree of agency that is not necessarily realistic. Due to decentralized energy distribution, a wind farm's operations involve thousands of disparate members of the public who may be totally unaware of their use of renewable sources of energy. In this sense, active "acceptance" is an imprecise term for the many forms of social relations that exist as renewables develop.
Although only a few renewable energy scholars have adopted the term "social license" (for instance, see Corscadden, Wile and Yiridoe 2012;Hall and Jeanneret 2015), several have described local opposition as serving the necessary political function implied by the use of the legal/political term "license." Barry and Ellis (2010) argued for new planning models that minimize the use of arbitrary authority to approve renewable energy generation facilities, and Ellis et al. (2009) noted that dissent serves an important purpose in democracies and that the critiques of renewable energy development can better inform societal decision-making. Thus, the social acceptance or social gap concept relates closely to social license through a shared emphasis on issues of legitimacy.

Exploring the relationship between social license and legitimacy
Despite their very different origins, all three varieties of social license invoke legitimacy. This is perhaps not surprising, since the concept of legitimacy "dates back to the dawn of organization theory" (Deephouse and Suchman 2008: 49; see also Deephouse et al. 2017;Suddaby, Bitektine and Haack 2017). Max Weber (1978) is typically credited with introducing the term into social theory, and for linking legitimacy with conformity to social norms and formal laws. In their classic work, Dowling and Pfeffer (1975: 124) distinguished three interdependent categories of organizational action-economic, legal, and legitimate-noting, "a legitimate purpose will not necessarily ensure resource allocation, nor will resource allocation necessarily ensure legitimacy." Similarly, although laws in a democratic society are likely correlated with societal norms and values, these correlations are apt to be partial, because: (a) changes in the two spheres may take place at different speeds; (b) contradictions are inherent in norms, whereas there is a greater presumption of consistency in legal frameworks; and (c) societies may tolerate certain activities without legally sanctioning them. Their model presciently foreshadows many elements of the three varieties of social license, especially the three strand model and the triangle model.
More recently, Scott (1995: 45) echoed these themes in his groundbreaking book: "Legitimacy is not a commodity to be possessed or exchanged but a condition reflecting cultural alignment, normative support, or consonance with relevant rules or laws." However, it was Suchman (1995: 574) who offered what has become perhaps the most widely cited definition of legitimacy: "a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions." Legitimacy is conferred on an organization when "stakeholders-that is, internal and external audiences affected by organizational outcomesendorse and support an organization's goals and activities" (Elsbach and Sutton 1992: 700). Stakeholders' legitimacy evaluation processes have been described as being an active judgment of the organization's observable properties and behaviours or a passive judgement relying on cues from others (regulators, media, or other opinion shapers), with reassessment loops (Bitektine and Haack 2015;Tost 2011). While organizations, projects, or products are most often the foci of legitimacy evaluations, stakeholders themselves also can be evaluated as to whether they have legitimate claims based upon "contract, exchange, legal title, legal right, moral right, at-risk status, or moral interest in the harms and benefits generated by company actions" (Agle, Mitchell and Sonnenfeld 1999: 508). Looking across these developments, one is struck by the extent to which legitimacy has become a widely studied and cited concept throughout the social sciences, one that entails a multi-directional evaluation by diverse stakeholders.
Along the way, multiple dimensions of legitimacy also have been articulated. Notably Aldrich and Fiol (1994: 648) differentiated cognitive legitimacy, or "the spread of knowledge about a new venture," from sociopolitical legitimacy, or "the process by which key stakeholders, the general public, key opinion leaders, or government officials accept a venture as appropriate and right, given existing norms and laws." Scott (1995) divided this latter dimension in two, resulting in three dimensions of legitimacy: regulative, normative, and cognitive, each linked with one of his three institutional pillars. In a similar vein, Suchman (1995) proposed a trichotomy comprised of cognitive, moral, and pragmatic legitimacy coupled with two temporal distinctions (episodic vs. continual) and two substantive foci (organizational actions vs. organizational essences), resulting in 12 distinct legitimacy types. Of relevance to issues of energy and resource development, Bansal and Clelland (2004: 94) proposed the concept of corporate environmental legitimacy, defined as the generalized perception or assumption that a firm's corporate environmental performance is desirable, proper, or appropriate.
Public administration scholars also have engaged with the concept of legitimacy, in terms of democratic accountability and governance. The emergence of social license mirrors a broader trend towards "networked governance," or a shift from traditional hierarchal and centralized governance to a more horizontal mode (Doberstein 2013). On this account, democratic accountability derives as much from judgments of the target population of policy initiatives, as much as from officials acting as the final decision-makers. Doberstein (2013: 604) termed this "social accountability" which again resonates with conceptions of "social license." Other public administration scholars have observed that legally constituted administrative bodies often issue guidelines that are not strictly "legal licenses" in the sense that they are not subject to ordinary mechanisms of legal review and democratic accountability (Houle and Sossin 2006). Relatedly, a recent definition of social license to operate put forward by business ethicists also stresses the role of legitimacy. Notably, Demuijnck and Fasterling (2016: 675-6) defined social license to operate as a "contractarian basis for the legitimacy of a company's specific activity or project. . . [which justifies] an institution or a moral or political rule by referring to the consent of all persons concerned with it." Looking across these literatures, what is striking is the interrelatedness of definitions of legitimacy and social license (see Table 1). While it may be tempting to declare that social license to operate is nothing more than a new name for the longstanding concept of legitimacy, there are useful distinctions. First, social license to operate is used to describe stakeholders' acceptance of a focal company or project to proceed with its activities. This tends to be a uni-directional evaluation of a company by its stakeholders. Conversely, the conferring of legitimacy can be multi-directional. While "legitimate" is often used to reflect stakeholders' evaluations of a company, it also can be the company's evaluations of its stakeholders, or stakeholders' evaluations of each other (for example, Mitchell, Agle and Wood 1997). Second, the pyramid model assumes that social license is built in successive layers: a company creates acceptance by following the rules, then approval by negotiating with stakeholders, and then identification through trustbuilding collaborations. Conversely, research into legitimacy building does not assume linear path-dependence: a company does not have regulative, then cognitive, and then moral legitimacy with its stakeholders. Legitimacy may be built in such a way, but not necessarily (Bitektine and Haack 2015). Legitimation tactics include ethos (credibility of the proponent), logos (logic of the argument), and pathos (emotional appeal) (Green, Li and Nohria 2009;Harmon, Green and Goodnight 2015). The credibility of authoritative endorsements (that is, media, regulators, influential others) can cause evaluators to passively rely on the judgments of these others, while logical inconsistencies or emotional surprises can cause evaluators to actively reassess the target (Tost 2011; see also Tormala, Petty and DeSensi 2010). Third, the three strand and triangle models of social license compartmentalize stakeholders according to their interests in the company or project. Sociopolitical/legal acceptance is a function of the company's technologies and policies meeting or exceeding regulatory requirements; community acceptance depends on demonstrating procedural justice, distributional justice, and trust; and market acceptance depends on formatting "Appraisal of action in terms of shared or common values in the context of the involvement of the action in the social system" (Parsons 1960: 175) Justification of organization's "right to exist" (Maurer 1971: 361) Implied congruence with the cultural environment, with "the norms of acceptable behavior in the larger social system" (Dowling and Pfeffer 1975: 122) Activities that are accepted and expected within a context are then said to be legitimate within that context (Pfeffer 1981: 4) An array of established cultural accounts that "provide explanations for existence" (Meyer and Scott 1983: 201) "Social fitness" (Oliver 1991: 160) A generalized perception of organizational actions as "desirable, proper or appropriate within some socially constructed system of norms, values, beliefs and definitions" (Suchman 1995: 574) "The endorsement of an organization by social actors" (Deephouse 1996: 1025) "Acceptance of the organization by its environment" (Kostova and Zaheer 1999: 64) "A social judgment of appropriateness, acceptance, and/or desirability" (Zimmerman and Zeitz 2002: 416) "The level of social acceptability bestowed upon a set of activities or actors" (Washington and Zajac 2005: 284) "The degree to which broader publics view a company's activities as socially acceptable and desirable because its practices comply with industry norms and broader societal expectations" (Rindova, Pollock and Hayward 2006: 55) "Having the approval, the broad acceptance of society to conduct its activities" (Joyce and Thomson 2000: 52) Meeting "social expectations . . . gaining support for the project from concerned groups, or stakeholders, over and above meeting any legal requirements" (BSR 2003: 4) The right of indigenous peoples and other affected parties "to participate in decision making and to give their free prior and informed consent throughout each phase of a project cycle" (Salim 2003: 21) "The demands on and expectations for a business enterprise that emerge from neighborhoods, environmental groups, community members, and other elements of the surrounding civil society" : 308) "The idea that industrial facilities must comply with tacit expectations of regulators, local communities, and the public in order to continue operations" (Howard-Grenville, Nash and Coglianese 2008: 77) "Ongoing approval within the local community and other stakeholders" Boutilier 2011: 1779) "Society's moral and political approval, sufficiently widespread and stable to allow legal approvals to proceed and to assure ongoing community support" (Cleland 2013: 1) "A form of social acceptance or approval. . . a socially constructed perception that your company or project has a legitimate place in the community" (Black 2013: 15) "Gaining, nurturing, and renewing legitimacy with local groups, stakeholders and communities" (Raufflet et al. 2013: 2229) "Ongoing acceptance or approval from the local community and other stakeholders" (Parsons, Lacey and Moffat 2014: 84) Note: Legitimacy definitions adapted from Bitektine (2011: 153). Social license definitions based on authors' analysis.
NEW FRONTIERS new technologies to fit within prevailing economic expectations. Conversely, how a company is evaluated along the multiple dimensions of legitimacy is not necessarily a function of who is doing the evaluating. Communities can and do judge a project according to multiple dimensions of legitimacy, such as: cognitive (Do I understand what the company is doing?), regulative (Are they following the law?), pragmatic (Does this project work for me?), and moral (Are they doing the "right" things?). We return to these distinctions in our implications section, to highlight how legitimacy research can address longstanding SLO debates in the Canadian context, such as: minority influence, the conversion of stakeholders into rightsholders, and (il)legitimacy spillovers. Given these differences despite similarities, one obvious question is: What is the nature of the relation between legitimacy and social license to operate?
Here we can only speculate, but it appears that legitimacy may provide organizations with a kind of "stock" or "surplus" to the extent its past efforts have created credibility, trust and so on. When an organization proposes a new project or initiative, this stock of legitimacy circumscribes or delimits the boundaries of such efforts. In this way, an organization's social license to operate appears to be a kind of liminal space in which new projects take shape. This space may be more or less constraining, or offer more or less latitude for action, depending on this stock of prior legitimacy, along with the variety of stakeholders and concerns which become involved (for example, First Nations, NGOs, governments, communities, and so forth). 7

Measuring social license
Although some have lamented that social license to operate is intangible , as with any social scientific concept, numerous measurement options exist. Of course, measurement depends on the definition of the construct in question. For instance, because the pyramid model assumes that "social license has levels" (Black 2013: 47), its proponents have attempted to demonstrate them empirically. In particular, the latest version of the pyramid model was reportedly inspired by Minera San Cristobal, a lead, silver, and zinc mine in Bolivia . The authors tracked changes in social capital over a 14-year period from 1994 to 2008 using "historical documents and the experience of persons who had been present throughout the life of the project using a basket of indicators and then verified through interviews with community members" Boutilier 2011: 1791; see also Thomson et al. 2010). The authors devised 24 statements aimed at measuring social license; however, "the subsets of statements meant to measure the separate layers of the social license to operate did not display the cumulative nature hypothesized by " (Boutilier and Thomson 2011: 3).
Considering these findings, in 2010 the authors revised the statements for use in Mexico and Australia. In 2011, they tested a revised set of 15 items, which according to their analysis loaded on four latent factors, dubbed economic legitimacy, socio-political legitimacy, interactional trust, and institutionalized trust. According to the authors, 85 per cent of cases conformed to a modified cumulative scale. For instance, stakeholders with a high score on the institutionalized trust factor "will always have high scores on all the other factors" (Boutilier and Thomson 2011: 4). Conversely, stakeholders with low scores on the socio-political legitimacy factor and the interactional trust factor "will never have high scores on institutionalized trust" (Boutilier and Thomson 2011: 4). Similarly, stakeholders with low scores on the economic legitimacy factor "will never have high scores on any of the other factors" (Boutilier and Thomson 2011: 4).
More recently, Black proposed measuring the pyramid model using 14 statements (see Table 2), which are rated on a five-point Likert scale, where 1 represents strong disagreement and 5 represents strong agreement. "To calculate a social license score, calculate the mean of each stakeholder's responses to the group of questions. If you have many stakeholders you could group them into categories that make sense, for example, local We are very satisfied with our relations with [name of company]. 4 [Name of company] does what is says it will do in its relations with our organisation. 5 The presence of [name of company] is a benefit to us. 6 [Name of company] listens to us. 7 In the long term [name of company] makes a contribution to the wellbeing of the entire region. 8 [Name of company] treats everyone fairly. 9 [Name of company] respects our way of doing things. 10 Our organisation and [name of company] have a similar vision for the future of this region. 11 [Name of company] gives more support to those who it negatively affects. 12 [Name of company] shares decision-making with us. 13 [Name of company] takes account of our interests. 14 [Name of company] openly shares information that is relevant to us. government, environmental action groups, and so on" (Black 2013: 51). Black recommended examining the standard deviation in the scores within the categories. "Wide variation suggests an 'average' score might be hiding some important differences between stakeholders that would be useful to know about for developing strategies" (Black 2013: 51). According to Black (2013), these social license measures have been used with around 5,000 stakeholders in 60 projects.
Relative to the pyramid model, several points are noteworthy. First, contrary to the hypothesized model, Boutilier and Thomson (2011) found no evidence for credibility as a separate factor between legitimacy and trust. Instead, they argued that socio-political legitimacy and interaction trust form a middle ground between mere economic legitimacy and institutionalized trust. Second, neither Boutilier andThomson (2011) nor Black (2013) provided any evidence for the reliability or validity of their specific questionnaire items or factor labels. Further work is needed to demonstrate whether the four factors they identified reliably and validly assess their proposed constructs. Third, the case study of the Minera San Cristobal mine introduces further conceptual slippage. Specifically, Boutilier (2011: 1792) claimed to measure "the ups and downs of social capital," but provided no justification for the subsequent conflation of social capital and social license. Fourth, Boutilier and Thomson (2011) interpreted their factors as cumulative. Namely, they argued that institutionalized trust depends upon the presence of the other three factors-or more generally, that legitimacy is a necessary but not sufficient condition for trust.
In a recent paper, two CISRO researchers investigated how the mining industry constructed the meaning of social license . They began by collecting 62 sustainability reports published by 19 companies between 2006 and 2009. Within these reports, they identified 133 mentions of the terms social license, social license to operate and license to operate. In 111 instances, agency was either implicit or absent, making it unclear who was acting or facilitating action. "This vagueness suggests that social license is essentially a metaphorical and rhetorical notion, bearing little resemblance to a license in the legal sense" (Parsons and Moffat 2014: 351). To further explore this possibility, they identified 121 instances where companies referred to acquiring (n 5 32), maintaining (n 5 83) or losing (n 5 6) social license. Interestingly, the authors noted: "No company discusses the prospect or consequences of failing to acquire a social license" (Parsons and Moffat 2014: 353).
Parsons and Moffat repeated their analysis using proceedings from a mining industry conference with the theme "sustaining our social license," and found 30 mentions of social license across 40 presentations and 35 abstracts. These mentions were distributed unevenly: mining companies represented 38 per cent of the speakers and 47 per cent of the mentions; governments represented eight per cent of the speakers and 20 per cent of the mentions; research institutions represented 18 per cent of the speakers and seven per cent of mentions; and indigenous organizations represented eight per cent of the speakers but none of the mentions. They concluded: "social license discourse has been adopted disproportionately within industry and government" (Parsons and Moffat 2014: 354). As with the sustainability reports, agency was implied or missing in 28 out of the 30 mentions, and only processes of acquiring (n 5 10) or maintaining (n 5 20) social license were discussed.
Whereas Boutilier, Black and Thomson (2012) discussed how mining companies might use the pyramid model for instrumental purposes (that is, acquiring and maintaining social license for their operations), Bice (2014) investigated the extent to which mining company conceptions of social license are (mis)aligned with the pyramid model. She studied the importance of social license from the companies' perspectives; whether they felt they had social license or not; and how this had changed, if at all, over time. Content and discourse analyses of 18 sustainability reports from five major Australian mining companies revealed that as a percentage of all issues discussed, explicit discussion of social issues had almost quadrupled between 2004 and 2007. Bice concluded that while social license is certainly important to the examined companies, how they go about determining whether or not they have it remains unclear.
In another CSIRO study, researchers focused on how mining companies engage with local communities, and how these processes affect the extent of their social license to operate (Moffat and Zhang 2014). Broadly speaking, Moffat and Zhang proposed that company practices influence trust, which in turn influences approval and acceptance. In other words, they reversed the relationship between legitimacy and trust assumed by the pyramid model such that trust is antecedent to acceptance and approval (legitimacy). To test their model, they conducted two internet surveys of local residents living in the areas affected by an Australian coal seam gas operation (CSG, also known as coalbed methane) who were not employed in the industry. They found that impacts on social infrastructure, contact quality and perceived procedural fairness are correlated with trust, which was in turn is correlated with acceptance.
Within Canada, a statistically representative survey of adults living in the western provinces of Alberta, British Columbia, Manitoba and Saskatchewan found that the farming industry had a net trust level of 62%, compared with net trust levels of 32% for forestry, 14% for mining and 13% for energy (Sajid 2014). 8 Lefsrud et al. (2015) analyzed 500,000 newspaper articles on Canadian resource development from 1989 to 2015. They found that "alarming" language (both highly negative and arousing) occurred twice as frequently in articles about oil than in articles about natural gas, forestry, mining, electricity, pipelines, and railroads. This pattern was exacerbated in the case of articles that mentioned "tar sands" as compared with articles that mentioned "oil sands." When viewed over time, they found that negative and arousing language associated with oil and gas articles had begun spilling over into discussions about mining, pipelines, electricity, forestry, and rail. Although corporations, regulators, and policymakers often discuss resource development in purely technical and economic terms, in doing so they fail to connect with increasingly emotional debates (Lefsrud et al. 2015).

Some implications and future directions
Developments such as fracking, pipelines, and even wind turbines are being increasingly debated online. Rather than a focal organization identifying its "stakeholders" based on its own interpretation of strategic or regulatory requirements (Freeman 1984;Mitchell, Agle and Wood 1997), individuals and organizations are increasingly able to selfidentify with an issue and opt into mediated and unmediated discussions. Social media dramatically expands such opportunities. Internetenabled technology (for example, web pages, blogs, YouTube, Wikipedia, Facebook, LinkedIn) and mobile apps (for example, Twitter, Snapchat, Flickr) allow virtual, unmediated conversations between quite distant actors. No longer do company managers, media reporters, or television producers solely determine who has the right to speak. User-generated social media provides massively distributed and open access to diverse media stages: anyone can distribute information, draw interest, criticize issues, and create global alliances.
Social media fundamentally challenges regulatory practices and consultation more generally. Previously, regulators and industries demonstrated "circular logic" whereby they invoked SLO in an effort to demonstrate alignment with stakeholder expectations. Claiming social license "can be used to mask the gap between company and stakeholder expectations. . .the industry's use of the term is interpreted here as an effort to disguise or silence opposition," even though companies, regulators, and their stakeholders "invariably have different expectations" (Owen and Kemp 2013: 31). Social media breaks this circle by giving previously dormant critics a forum to publicly voice their differing expectations to a global audience, without resorting to traditional legal, political, or media channels. Particularly compelling messages can even travel between issue fields and change adjacent organizations and institutions (Harmon, Green and Goodnight 2015), giving minorities the ability to influence and even change majority opinions through social media channels (Gardikiotis 2011;Moscovici 1976).
But this raises several pressing questions related to the possibility of a vocal minority overruling the will of a less vocal majority (for example, Simpson 2014). 9 For instance, whereas research on environmental justice and social movements has emphasized the need to give voice to those affected by industrial facilities and large-scale energy projects, the concept of social license to operate suggests an opposite dilemma whereby small but vocal stakeholder groups may be able to thwart activities that are not seen as problematic by most stakeholders. Recalling Pfeffer and Salancik's (1978) classic treatise, the issue appears to depend on whether the focal organization has access to substitute sources of critical resources.
Another question regards the mechanisms underlying such processes. Here, it may be worthwhile to consider how visual and emotional appeals are particularly effective at capturing attention, re-problematizing taken-for-granted understandings, allowing alternative interpretations of ambiguous texts, and creating openings for contestation (Jones et al. 2017;Lefsrud, Graves and Phillips 2017;Meyer et al. 2013;Voronov and Vince 2012). Indeed, challengers use visual rhetoric to create powerful images, for instance, of death and sanctity, which effectively critique a company's moral legitimacy and override any counter-arguments of regulatory or pragmatic legitimacy. As a result, the moral concerns of environmental groups, global publics, and foreign governments become more influential than the instrumental concerns of local employees or regulatory evaluations of provincial or national governments (Bitektine 2011;Tost 2011).
Third, it is important to understand how the implications of social license to operate are evolving in the context of Indigenous engagement in regulatory processes and policy-making. On the one hand scholars have pointed out that First Nations are rightsholders who legally must be consulted (Colton et al. 2016). Thus, Indigenous communities already hold a legal status in regulatory and permitting processes that exists independent of concerns related to an organization's social license to operate. Moreover, the concept of social license to operate has created new space for dialogue and potentially for shared decision-making among Indigenous communities, affected publics, and the Crown (Doberstein 2013). Recent scholarship has suggested that multi-level governance is a new form of Aboriginal policy-making that allows "Indigenous governments and organizations [to] participate as co-producers of public goods (for example, policy and political decisions)" (Alcantara and Spicer 2016: 188). One example that may be worth exploring in detail in future research is how the emergence of the concept of a social license to operate inspired the National Energy Board (NEB) to formalize a requirement for Indigenous expertise within its membership and in regulatory decisionmaking (Natural Resources Canada 2016).
In sum, interested actors have an amplified voice regarding proposed developments, even those from which they were previously excluded. This also suggests that for researchers to understand how social, political, and legal licenses are withheld, created, maintained, or withdrawn, they must consider virtual (for example, Twitter) and episodic forums (for example, multi-level governance and decision making), in addition to traditional, physical forums (for example, administrative tribunals, courts of law).

Conclusion
In many societies, social license to speak plainly has been accorded a privileged role in challenging and realigning legal, political, and economic affairs. With mounting criticism of energy developments, companies and regulators have been increasingly invoking the concept of SLO, usually as a means of legitimating their actions, regardless of underlying sentiment. In this article, we reviewed different models of social license, and showed how it appears to be little more than a new name for legitimacy. By further discussing some implications for stakeholder engagement and evolving models of regulation, our aim has been to motivate further research into the theorization, operationalization, and empirical functioning of social license to operate.
Notes a new or under-studied phenomena. However, because such samples are not representative, they cannot be generalized to a larger population. 5 The construal of legitimacy as merely a normative concept represents an incomplete understanding relative to prior academic formulations. Rather, legitimacy has been conceptualized as multifaceted, entailing not only a normative dimension, but others as well (for example, cognitive, socio-political). 6 Implicit in such a proposition is the assumption of a more complex multilevel relationship, an implication that to our knowledge has not been fully explored in prior work. 7 We thank the editor, Evert Lindquist, for inspiring our discussion in the preceding paragraph. 8 Net trust was measured as the number reporting trust minus the number reporting distrust. 9 We thank an anonymous reviewer for sparking discussion of this issues.