Deliberating with the Autocrats? A Case Study on the Limitations and Potential of Political CSR in a Non-Democratic Context

Extant literature on Political CSR and the role of governments in the governance of business conduct tends to neglect key implications of the political-institutional macro-context for public deliberation. Contextual assumptions often remain rather implicit, leading to the need for a more nuanced, explicit and context-sensitive exploration of the theoretical and practical boundary conditions of Political CSR. In non-democratic political-institutional contexts, political pluralism and participation are limited, and governmental agencies continue to play the most central role in regulation and its enforcement. Drawing on a qualitative case study on a nuclear energy project in the Hashemite Kingdom of Jordan, we show how both governments and MNCs co-create a context hostile to socially responsible business conduct in the sense of Political CSR. Utilizing rich qualitative data derived from interviews, public documents, and participant observation between 2008 and 2018, we specifically illuminate how—through which interactions and strategies—the multiple governmental and corporate actors involved counter civil society demands for public deliberation, indicating the limitations of Political CSR in non-democratic political-institutional contexts. We particularly contribute to the development of Political CSR by analysing the role of coercive and discursive forms of power. We thus offer a more nuanced perspective on the role of governments in constraining the room for public deliberation in the sense of Political CSR.


Introduction
A prominent approach to the responsible management of international business is the concept of Political Corporate Social Responsibility (Political CSR) (Scherer & Palazzo, 2007, 2011. Research associated with Political CSR normatively and empirically discusses how corporations increasingly assume "political coresponsibility" (Scherer & Palazzo, 2007: 1109 by closing governance gaps through (self-)regulation or providing public goods when a national government fails to do so (Matten & Crane, 2005;Scherer & Palazzo, 2007, 2011Scherer et al., 2016). They mainly do so by "engaging in public deliberations, collective decisions, and the provision of public goods or the restriction of public bads in cases where public authorities are unable or unwilling to fulfil this role" (Scherer et al., 2016: 276). From a discourse ethics and deliberative democracy perspective, decisions of public interest need to be legitimized through public deliberation (see, e.g., Habermas, 1996). It is crucial for the corporation concerned, and particularly Multinational Corporations (MNCs) as powerful actors, to obtain moral legitimacy (Suchman, 1995), which is "socially constructed by giving and considering reasons to justify certain actions, practices, or institutions" (Palazzo & Scherer, 2006: 73). The (deliberative) quality of a decision-making process then determines the legitimacy of a political decision (Gilbert & Behnam, 2009: 225), and holds the potential to solve conflicts between different stakeholders.
Extant research rather narrowly focuses on "the standard case of MNCs from 'western' countries with decent democratic institutions and rule of law regimes at home that operate in fragile states" (Scherer et al., 2016: 285). MNCs, however, operate in a variety of political-institutional contexts Political CSR does not sufficiently take into account yet (Westermann-Behaylo et al., 2015). Moreover, we find 1 3 a differentiation of political-institutional contexts particularly relevant for an analysis of the boundary conditions of Political CSR. The need for a greater sensitivity towards the nuances and variations of political-institutional contexts of CSR and their respective power disparities (Gond & Nyberg, 2017) is also emphasized in recent research that adresses the implications of often under-explored political institutions and geographical regions for CSR (e.g., Jamali et al., 2020). We seek to examine the concrete ways in which nondemocratic political-institutional contexts with their distinct power asymmetries present obstacles to corporate engagement in public deliberations with civil society actors.
Generally, non-democratic governments "attempt to coopt, subdue, or eliminate all sources of political power outside the state system, whether they are mass based or controlled by social elites" (Perlmutter, 1981: 26). They actively limit pluralism and political participation (Linz, 2000). As the 2022 Democracy Index of the Economist Intelligence Unit illustrates, studies related to the politicalinstitutional conditions for doing business in authoritarian or non-democratic contexts continue to become more relevant, as there have been global regressions of democracy, both in developing country contexts and historical strongholds of democracy (The Economist Intelligence Unit, 2022).
We use a qualitative case study on a nuclear energy project involving a multitude of both domestic and foreign government agencies and corporations in the Hashemite Kingdom of Jordan to develop the theoretical scope of Political CSR in this under-explored, non-democratic political-institutional context. The nuclear energy project was launched in 2008 and comprised the construction of nuclear power plants (NPPs), a research and training reactor, and uranium exploration. It has since undergone shifts in terms of scale and scope, involving changes of project partners for the construction of the NPPs in 2018 (Global Construction Review, 13 June, 2018). We focussed our data collection on the period 2008-2018, as this time reflects the height of contestation through several societal actors, e.g. environmental activists (Su, 2013). A parliamentary vote to suspend the nuclear project in 2012 was ignored by the Jordanian government, which moved forward with the project, and said that it would only discuss the project with experts that were qualified (Seeley, 2014). Moreover, we experienced how interview partners opposing the project were harassed and threatened by security services during field research. Drawing on our rich field data, we elucidate through which interactions and strategies governments and MNCs co-create a context hostile to socially responsible business conduct in the sense of Political CSR. Through this enquiry, we seek to answer our overarching research question: What are the implications of a non-democratic political-institutional context and its distinct power asymmetries for the Political CSR of MNCs? More specifically, we explore how the non-democratic context under study poses challenges to the corporate engagement in public deliberation.
To explore the boundary conditions of Political CSR in non-democratic contexts, Jordan is a particularly interesting case of a non-democratic state. According to recent integrated accounts of the Varieties of Capitalism and National Business System literatures, Jordan can be classified as a developmental state (Fainshmidt et al., 2018). At the same time, it displays features of a closed anocracy, i.e. a political system that relies on authoritarian institutions whilst including single democratic features such as elections (Marshall & Elzinga-Marshall, 2017). Whilst most pertinent political indexes do not classify Jordan as a fully consolidated autocracy like Saudi Arabia, there are signs of progressive autocratization. For example, Freedom House has recently downgraded Jordan from "partly free" to "not free" (Freedom House, 2021), indicating an authoritarian shift that is also noticeable in other indexes (e.g., V-Dem Institute, 2021). This authoritarian shift involves increasing crackdowns on political speech and popular protests (Silva, 24 March, 2021) and is also reflected by the arrest of one of our interview partners. In the dynamic context of the "Arab Uprisings", it presents a particularly relevant political-institutional context for Political CSR. Great parts of the Jordanian population thus increasingly expect to be involved in political decisions, whilst protests (e.g. large-scale protests of public school teachers in July 2020) are met with harsh crackdowns involving arrests, threats and violence, often under reference to emergency laws (Nusairat, 09 April, 2021).
We intend to make three contributions. First, we seek to contribute to the further development of Political CSR through contextualization. We provide empirical research on this under-researched, yet particularly relevant politicalinstitutional context. We specifically show how in a context differing from the standard cases of Political CSR, the arguably most central provision of corporate engagement in public deliberation faces serious institutional constraints that fundamentally relate to power asymmetries characteristic of non-democratic contexts. Second, we provide starting points for more context-sensitive analyses through the deconstruction of central contextual assumptions dominant in the literature, and third, we offer a more nuanced perspective on the role of governments and power asymmetries in constraining the room for maneuver for Political CSR by discussing the particularities of non-democratic regime types. This also allows us to contribute to the growing body of literature on the role of governments in the (responsible and irresponsible) governance of business conduct (Kourula et al., 2019). Beyond addressing some of the "descriptive inaccuracies" (Whelan, 2012: 717) of Political CSR and answering Scherer et al.'s calls for an expanded research agenda (Scherer et al., 2016), we expand its theoretical scope by exploring the theoretical and managerial implications of a non-democratic political-institutional context and the underlying structures and dynamics of power.
The article proceeds as follows. First, and focussing on its normative calls for corporate engagement in public deliberation, we discuss key assumptions of Political CSR on political-institutional context, highlighting the role of power in public deliberation. We subsequently discuss the concept's limitations in under-researched non-democratic politicalinstitutional contexts. The methodology section outlines the research context and empirical setting whilst describing our processes of data collection and analysis. We present and discuss the findings in the light of theory. Finally, the conclusion suggests areas of further research.

Political CSR in a Non-Democratic Context
In the following sections, we problematize key assumptions of Political CSR scholarship on political-institutional context. We discuss how the extant literature problematizes the prevalent underestimation of power and the effects of power asymmetries on public deliberation. Finally, we explain how power asymmetries that are characteristic of non-democratic contexts constitute boundary conditions for Political CSR in contexts deviating from its standard cases.

Political CSR: Public Deliberation and Power
Political CSR is a normative concept that assigns corporations the responsibility to engage in public deliberation, self-regulation and public good provision (Scherer & Palazzo, 2007, 2011Scherer et al., 2016). As "a significant part of global production has been shifted to locations that lack democratic control and where there is no rule of law" (Scherer & Palazzo, 2011: 902), corporations can no longer solely rely on national governance through the national state. Building on elements of Habermas' concept of deliberative democracy (1996), Political CSR requires corporations to take part in processes of democratic will-formation as politics already starts "at the level of deliberating civil society associations" (Scherer & Palazzo, 2011: 908). Enacting laws and producing regulation is no longer a preserve of state authorities, and the state, civil society and corporations jointly contribute to governance. MNCs thus turn into political actors who co-create soft law and governance regimes, e.g., agreements, principles and declarations that are not legally binding (Scherer & Palazzo, 2011: 907).
The corporate engagement in public deliberations serves as the modus operandi in which MNCs are supposed to engage in self-regulation or public good provision. Not only is corporate legitimacy created through deliberation, but national governance systems are partly replaced by new, heterarchic forms of governance such as Multi-Stakeholder Initiatives (MSIs) that generate soft law and sustainable industry standards together with state-and non-state actors (e.g., Huber & Schormair, 2019). (Self-)regulation then emerges from communicative processes that ideally are designed to ensure social acceptance, and to enhance corporate legitimacy and accountability (Palazzo & Scherer, 2006;Gilbert & Behnam, 2009;Scherer et al., 2013). Following this approach, powerful MNCs in particular need to participate in public deliberation to acquire moral legitimacy, which "reflects a prosocial logic that differs fundamentally from narrow self-interest" (Suchman, 1995: 579). This engagement ensures that stakeholders perceive an MNC and its actions as legitimate based on an exchange of reasonable arguments (Palazzo & Scherer, 2006: 73). Political CSR therefore advocates discursive politics rather than power politics, and replaces liberal democracy with deliberative democracy (Scherer & Palazzo, 2011).
Such deliberative approaches to governance and selfregulation require a context in which the outcome of deliberation is not primarily based on the power of the respective actor, but the much-quoted unforced force of the better argument Habermas envisions in "Between Facts and Norms" (1996), where he also addresses the general tension between power and (rational) communication. There is an ongoing debate in the interdisciplinary deliberative democracy literature that discusses this tension, arguing that both power and self-interest were generally underestimated categories in deliberative democracy approaches (Pellizzoni, 2001;Mansbridge et al., 2010). Conceptually resolving these tensions is beyond the scope of our paper. We address them to the extent they are reflected in more critical accounts of Political CSR that highlight the underestimation of the role of power in deliberative governance efforts (Banerjee, 2008(Banerjee, , 2010Moog et al., 2015).
Power is a central category in our analysis. The literature offers various understandings of power (Kohn, 2000;Kadlec & Friedman, 2007). On a basic level, we define power as the inherently relational capacity to cause or change outcomes, either positively or negatively from the perspective of the respective agents involved (Lukes, 2005). This capacity can be materialized through coercive power (e.g. through physical repression) (Haugaard, 2014), or discursive power (e.g. shaping norms and ideas through communication) (e.g., Banerjee, 2018). The notion of discursive power is particularly relevant for further developing Political CSR and can be further specified by drawing on Lukes' "Three Faces of Power" (2005). Luke distinguishes between decision-making power, agenda-setting power, and ideological power, all of which have a clear discursive dimension. Decision-making power refers to the power of governments to make legitimate decisions on behalf of the people. It includes open and observable conflicts of interests in policy debates, which actors are fully aware of. A key question in this context is: who holds the power to determine what is up for such debate? The second face of power addresses this question by highlighting how powerful actors can keep concerns of comparatively less powerful actors off the agenda. Agenda setting power then materializes through non-decision making and the exclusion of certain issues from public deliberation. The third face of power concerns a more latent conflict: comparatively less powerful actors are not even aware of their real interests as these had been shaped through ideological power: "Shaping their perceptions, conceptions and preferences in such a way that they accept their role in the existing order of things" (Lukes, 2005: 24). Women supporting a patriarchal society would be a classic example.
We find these distinctions useful as they highlight the need to account for the more latent dimensions of power. Whilst we take coercive, repressive forms of power into account (e.g., crackdowns on political opposition or extensive surveillance), we are thus particularly interested in the role of discursive power exerted by a non-democratic government. We accordingly focus on how decision-making power, agenda-setting power and ideological power materialize in this case study.
Power asymmetries are observable in all political contexts, and bear the risk that strategic considerations of corporate actors dominate deliberative efforts through their discursive power (Banerjee, 2018), which enables them to operate with a notion of legitimacy even in challenging environments like non-democracies. Participants in public deliberation differ from each other with regard to their social power, and hence differ in their power in public deliberation (Allen, 2012). Sabadoz and Singer (2017: 188), for example, caution proponents of deliberative approaches not to overestimate "the power of civil society to hold firms to account". Agonistic pluralist thus highlight the need to take dissensus and power imbalances seriously, and propose arbitration as a way to reduce "power asymmetries between the involved parties while respecting them as legitimate adversaries" (Dawkins, 2015: 17). In the following section, we deconstruct key contextual assumptions of Political CSR and its critics by discussing the distinct power asymmetries found in non-democratic contexts.

The Need for Contextualization: Political CSR and Power in Non-Democratic Contexts
Political CSR scholarship predominantly builds upon particular contextual assumptions. It is mainly proposed as a response to fragile statehood and eroding state power (see, e.g., Matten & Crane, 2005;Scherer & Palazzo, 2007, 2011Scherer et al., 2016). A potential problem resulting from that is an under-or over-estimation of obstacles to socially responsible business conduct in contexts differing from the literature's standard cases. The persistence of state actors in the governance of business conduct, however, has started to become a central theme in scholarly debates on (Political) CSR (see, e.g., Moon et al., 2010;Gond et al., 2011;Knudsen & Brown, 2015).
Governments continue to intervene in business activities in multiple ways (Peterman et al., 2014) and thus shape governance spaces (Kourula et al., 2019) for MNCs and other non-state actors-e.g., the room for maneuver regarding socially responsible business practices. This role of governments in (Political) CSR in general and deliberative approaches in particular is often discussed based on rather implicit assumptions that often do not differentiate democratic and non-democratic political-institutional contexts. The same is true for the recent debates on the role of power asymmetries in (Political) CSR outlined in the previous section. State power is thus often analytically linked with the observation of human rights violations, or the ability of autocracies like China to resist public opposition and pressure (Santoro, 2015;Schrempf-Stirling, 2018). Nondemocratic governments might utilize CSR for their own governance agenda, whilst "acting as the main mediator of societal demands to business" (Hofman et al., 2017: 12). A deeper understanding of such under-explored non-democratic contexts and their particular power asymmetries is an analytical prerequisite for exploring the contextual boundary conditions of Political CSR.
Power asymmetries in non-democratic contexts differ from those in democracies in the following ways. Pluralism and political participation are generally actively limited in non-democratic states, whilst the executive lacks constitutional accountability (Linz, 2000). The security apparatus plays a central role in the coercion and control of the population, and political decisions are commonly made centrally by a small elite (Sassoon, 2016). Where political parties exist (as is the case in Jordan), they mainly serve as instruments of authoritarian co-optation through "hierarchical assignment of service and benefits, political control over appointments, and selective recruitment and repression" (Svolik, 2012: 163). This illustrates that non-democratic governments do not rely on physical repression alone. They also need to assert their discursive power by maintaining their hegemony in public deliberation. This often happens through authoritarian deliberation, which refers to government-controlled forms of political participation that do not result in regime democratization or political empowerment of citizens (He & Warren, 2011).
In this context, the prospects for the inclusion of civil society actors like local communities or NGOs into corporate democratic decision-making are limited. The repressive context found both in fully consolidated autocracies and more open or transitioning non-democratic states like Jordan then present a political-institutional context setting considerable boundaries for deliberative processes that include civil society actors. Extant research has rarely addressed howthrough which discursive strategies-societal demands for public deliberation are countered in such contexts. During data collection and analysis, we were thus particularly interested in whether and how non-democratic power dynamics materialize through discourses surrounding the nuclear energy project.

Qualitative Case Study on a Nuclear Energy Project in a Non-Democratic Context
With our case study on a nuclear energy project in Jordan involving a multitude of governmental and corporate actors, we contribute to a more nuanced understanding of how power asymmetries affect the prospects of public deliberation in non-democratic contexts. Jordan qualifies as a "closed anocracy", which means that the prevalent mode of authoritarian decision-making is complemented through single democratic elements such as elections (Marshall et al., 2019). When it comes to economic activity, the emphasis lies on hierarchical coordination, as well as the monarchy's need for economic "development" (Fainshmidt et al., 2018). As described in the introduction, Jordan has recently experienced an authoritarian shift and increasingly repressive responses to popular protests and political opposition (Nusairat, 09 April, 2021). In political contexts where authoritarian politics prevail, political power is exercised by a leader or a small group "within formally ill-defined limits" (Linz, 2000: 159). This also affects negotiations over "which projects the state should pursue, and, by implication, which individuals will benefit from the translation of political influence into economic power" (Brownlee, 2007: 212), and shapes how (responsible) business takes place.
However, the concept of public deliberation is not completely absent in non-democratic systems of governance. The "Arab Uprisings" have shown that issues related to the distribution of power or the interlinkages of business firms and the state are not only debated by a variety of stakeholders, but also render authoritarian politics more dynamic than they seem at first sight. Understanding this background is imperative when aiming at understanding the possible roles MNCs can play in this context.
Jordan imports roughly 97% of its energy needs, which amounts to spending around 20% of its GDP on energy imports (The World Bank, 2016). Given its lack or early stage of exploration and development of significant energy resources (EIA, 2018) as well as disruptions in affordable gas imports from Egypt combined with a rising domestic energy demand, Jordan suffers severe energy insecurity. A crucial part of the nuclear energy program is the Jordan Research and Training Reactor (JRTR) based at the Jordan University of Science and Technology (JUST), close to the border to Syria. In 2009, a South Korean consortium including private and public actors was selected to build the JRTR. The consortium comprises of the Korean Atomic Energy Research Institute (KAERI), the government-owned Korea Electric Power Corporation's subsidiary KEPCO E&C, and DAEWOO Engineering & Construction, a stock company engaged in large-scale projects operating in Asia, Africa, the Middle East, Europe and the U.S. (DAE-WOO E&C, 2016a). Russia's national nuclear corporation Rosatom won the contract as Jordan's strategic partner, who would both invest in and operate nuclear power plants consisting of two 1000 Megawatt (MWe) reactors within a Build-Own-Operate framework. In late 2018, however, the Jordanian government made the final decision not to implement the large nuclear reactor project with Russia, citing project financing concerns, and rather explore the option of small modular reactors (SMR) instead (Global Construction Review, 13 June, 2018). Russia stays in the frame regarding the development of SMR, whilst JAEC has entered negotiations with Chinese firms on the feasibility of a large nuclear reactor (Jordan Atomic Energy Commission, 2021).
According to our interviews, newspaper articles and reports of nongovernmental organizations (e.g., Environmental Justice Atlas, 2018), stakeholders embedded in the local culture (e.g., residents of the sites of the nuclear reactors, environmental activists) commonly describe the nuclear energy project as irresponsible and attribute a series of highly contested issues to it. These issues range from mismanagement and corruption affecting nuclear safety and the ignorance of industry standards to a lack of stakeholder involvement and repression. We do not evaluate whether these claims and attributions are fully correct or whether the Jordanian government should pursue nuclear energy or not, especially given Jordan's severe energy insecurity. Rather, we are interested in how the non-democratic context in question affects public deliberation in the sense of Political CSR from the perspective of local civil society.
Local residents in the proposed area of the NPPs thus threatened a "war of resistance" against the project, vowing to prevent construction crews from entering the area (Luck, 2013). Jordanian members of parliament had voted to suspend the project in 2012 (Magid, 2016). The Jordanian government nevertheless decided to continue with project implementation. Parliamentarians and environmentalists had also repeatedly accused the Jordan Atomic Energy Commission (JAEC) of misleading the public regarding the economic feasibiliy of the project by presenting too optimistic data and omitting central information on project costs (Omari, 30 May, 2012). Another accusation relates to the governmental decision in favor of project implementation prior to any feasibility studies (Luck & Omari, 2012). The number of public protests, demonstrations and calls for public deliberation on the project has since grown (Namrouqa, 2014).

Methodology
We followed an exploratory, interpretive approach (Gioia & Pitre, 1990: 588;Suddaby, 2006) and sought to remain open to potentially surprising empirical and theoretical puzzles (Gioia et al., 2013). Epistemologically, we recognize that research is always at least partly subjective (Alvesson et al., 2008;Sandberg & Alvesson, 2021). Accordingly, we generally understand qualitative research as a "subjective methodological approach that acknowledges that the world as we know it is the (partial) result of our own social constructions, including those of the researcher" (Whiteman & Cooper, 2016: 123).

Case Selection
We selected the case of Jordan's nuclear energy project as it is, first, situated in the under-studied context of interest that oscillates between autocracy and anocracy. Second, it presents a situation where local civil society calls for public deliberation. Third, a multitude of public and private actors are involved, whose interactions contribute to the phenomena observed. Whilst we do focus on these actors, more have been involved in the project, e.g. the International Atomic Energy Agency (IAEA), European MNCs (e.g. French Areva), and multiple international consultancy firms.

Data Collection
For the purpose of clarity, we refer to data collection and data analysis sequentially, whilst in reality, we moved iteratively between them. This study is based on two core principles: constant comparison and theoretical sampling. Regarding theoretical sampling, we collected a variety of data from different sources, and let the research process guide further data collection (Glaser & Strauss, 1967;Corbin & Strauss, 2008). We sent 42 interview requests to the central corporations involved in the project, i.e. Russian Rosatom and its subsidiaries, Korean DAEWOO E&C and KEPCO, as well as French AREVA between 2014 and 2018. We used multiple channels, such as official company contact forms or emails, and directly approached current and former employees on platforms such as LinkedIn. We did not receive a single response, highlighting the challenges of data collection in this research context.
When interviewing civil society actors, we noted that hardly ever did opponents of the project refer to the corporations involved as dominant actors they would direct their allegations to. Rather, they mostly directed their concerns to the Jordanian government. We therefore concentrated our efforts on acquiring interviewees from Jordanian governmental agencies involved. They, however, were equally unresponsive as the corporate side, with the exception of one member of the Jordanian Senate whom one of us had met personally in another context. Nevertheless, we were able to interview three current and three former Jordanian government officials. Two of the current and all of the former government officials had directly been involved in the nuclear energy project in leading positions at JAEC and a nuclear oversight body.
Furthermore, we collected data on the governmental perspective on the project from Jordanian state media (119 articles between 2012 and 2016) and diplomatic cables published by Wikileaks (52 cables between 2003 and 2010), both of which provided for insights into the inner workings of governmental politics and reasoning. We also included 21 Russian state media articles and 13 Russian government publications into the database. The only other governmental representative willing to talk to us was a senior Korean diplomat with a personal interest in sustainability. In addition to the 16 interviews informing our analysis and the 1.5 h of direct participant observation of a tribal assembly in the Northern desert intended for a discussion of the nuclear energy project, we coded 264 documents covering a timeframe from 2003 to 2018, with a focus on the period of 2008-2018. Table 1 provides a summary of the data sources used.
The interviews lasted between 30 min and 2 h and were conducted by one of the authors in English and Arabic with the help of a local translator. Some involved spending hours in the home of the interview partner, e.g. with a tribal leader in the area of the proposed power reactor site or a representative of a local community in the area of the JRTR. Given the sensitive nature of the topic, the interviews could not be recorded. Instead, the interviewer wrote detailed memos and protocols, and we followed up on some of the interviews to verify whether a theme that seemed to emerge from the data would be confirmed in another, more informal conversation. Areas of inquiry covered perspectives on the project's overall rationale, challenges and critical issues, experiences related to project planning and implementation, and assessment of interactions and behavior of the project's central actors. Details on the interviews and the participant observation are provided in Table 2.
Our interviewees were careful speaking to a "Western" researcher on a very sensitive topic given the omnipresence of state security and the lack of civil liberties and political rights (Morgenbesser & Weiss, 2018), which highlights the role of powerful actors in our research process. Criticism of the King, for example, would have to be avoided in the interviews, as it might have resulted in criminal prosecution. As opponents of the nuclear energy project were more responsive to interview requests, the first round of interviews was dominated by very critical, often emotional, accounts of the situation. Interviews taking place in Amman were usually conducted in busy and loud hotel lobbies or cafés to avoid recordings through security personnel, and involved a longer phase of trustbuilding, especially when interviewing opponents of the nuclear energy project.

Data Analysis
As described, the analytical process started with first thinking about how one could possibly obtain sufficient data on such a rather sensitive topic. In terms of coding, it proceeded in four stages of constant comparison (Glaser & Strauss, 1967). To support the initial organization and systematization of data, we importet all data (except 119 Jordanian state media articles) into the software MAXQDA. The data were initially coded by one author. To ensure a certain degree of intercoder-reliability (O'Connor & Joffe, 2020), however, we frequently revisited our data together during the writing process, discussing emerging themes in light of the literature.
First order codes mainly consisted of in-vivo, descriptive, process and causation codes. The goal of open coding was to identify statements, themes and dominant issues hinting at processes related to the repression of public discourse on the project. In the next step, using axial coding, we looked for relations between the first order codes, grouping them into broader themes (Corbin & Strauss, 2008;Gioia et al., 2013), specifically focusing on contextual circumstances and interactions. Constantly comparing these themes and the data, we consulted political science literature on nondemocratic contexts and energy as well as geo-politics, especially to illuminate those areas of inquiry we were not able to cover via interviews (e.g. with regard to Russian government agencies and their intertwinement with corporations). Since we found it particularly hard to understand the role of the Korean consortium, we conducted an expert interview and asked further country experts for advice regarding literature. Additionally, we conducted an interview with a Korean diplomat, which proved particularly insightful especially in comparison with equally analyzed media articles, company publications and the expert interviews.
Continuing our iterative process of data analysis, we then consulted further pertinent political science literature and went through the first order codes we had derived again to construct the emerging theoretical relationships among the second order themes, finally collapsing them into aggregate dimensions. The focus of data analysis was on interactive processes and their consequences with regard to deliberation whilst being sensitive to context (Corbin & Strauss, 2008: 114).
Power asymmetries affected both data collection and data analysis. We were careful to critically and continuously reflect upon our own positionality throughout the whole research process (Abdelnour & Abu Moghli, 2021). This politically reflexive approach inter alia resulted in not conducting some of the interviews we could have conducted, or not using some of the information we obtained in our data analysis, in order to avoid potential harm to our research partners. For example, we had to cancel a second research trip to Jordan when a powerful business figure with close ties to the Jordanian government one of us knew personally injected himself into the research project. After having heard of the interview with the member of the Jordanian Senate, he independently scheduled interviews with high-ranking decision-makers within the nuclear energy program. We had consciously avoided contacting this particular group of government officials because of their distinctive (coercive) power. We found them too powerful and influential for us to be able to guarantee that no negative repercussions for our research partners and ourselves would result from such an interview. He also insisted of having his driver transport the interviewer from interview to interview, and asked for a list of all of our interview partners, which we were not willing or authorized to comply with. This illustrates the politics of access and the role of power in research in challenging contexts: Access to powerful people may promise deep and interesting insights, but can also be withdrawn or become overall problematic (Cunliffe & Alcadipani, 2016). Finally, experiencing this potential of harm to research partners and recognizing the resulting responsibility in the steps of data collection and analysis creates tensions that required constant critical self-reflection. Figure 1 provides an upfront summary of how the Jordanian government exercises its coercive (dotted lines) and discursive power (continuous lines) through strategies designed to effectively counter civil society demands for public deliberation. As briefly described in the empirical setting above, civil society actors such as tribal representatives or environmentalists have intensified their calls for genuine stakeholder involvement and public deliberation on controversial issues surrounding the nuclear energy project. We found that the most dominant and visible actors are indeed the different governments involved, who, through their interactions, shape a context inhospitable to responsible business conduct in the sense of Political CSR. Especially from the perspective of local civil society, the Jordanian government is the most dominant and visible actor (highlighted in black). The second most dominant and visible actor is the Russian government, together with state-owned Rosatom (highlighted in dark grey). The least visible actor is the Korean project partner (highlighted in light grey). Whilst the Jordanian government employs a range of sophisticated strategies for countering the demands of civil society for public deliberation, the Russian and Korean project partners focus on evading public deliberation.

Not Just Repression: The Role of the Jordanian Government in Public Deliberation
Our findings show that from the perspective of local civil society, the Jordanian government employs five major, partly interlinking, strategies to counter public resistance to project implementation: (1) Repressing public deliberation, (2) distorting public deliberation, (3) delegitimizing opponents and critics, (4) connecting the project to grievances, and (5) creating time pressure. Violent repression of public deliberation as a form of coercive power is just one means the Jordanian government employs. The other four strategies found in our data represent forms of discursive power. Among those four strategies, the most dominant strategy used by the Jordanian government, can be described as distorting public deliberation (highlighted in grey), as we explain in detail below (Fig. 2).
1. Repressing public deliberation. Our interview partners opposing the nuclear energy project frequently referred to intimidation through the state's security agencies and escalating violence. They explained how they were under increasingly tight surveillance, intimidated or arrested, and allegedly subjected to torture (interviews C1_LC_2014, C2_E_2014, C5_E_2014). One of our interview partners, for example, regularly had to report to the domestic secret service, and has recently been arrested for spreading "false information that leads to concern and affects a public institution" regarding safety aspects of the nuclear energy plants (source left blank for identification potential). Further measures include the (allegedly) retaliatory sabotage of the opponents' business activities (interview C5_E_2014). During field research, we also encountered intimidatory forms of surveillance. When attending a tribal assembly, an interview partner who had taken the interviewer there pointed to a man he identified as a member of the mukhabarat, the Jordanian domestic secret service. This is also in line with the general political development of increasing repression in Jordan (expert interview 2), which seems to be based on the understanding that "free political discussion is, in itself, a threat to national security" (Jarrah, 2009). Based on these findings, the strategy of repressing public deliberation applied by the Jordanian government can be described as an expression of coercive power.
2. Distorting public deliberation. Even non-democratic states seek "some appearance of accountability" (Tombs & Whyte, 2003: 225). Accordingly, they strive for ensuring the dominance of the governmental narrative in public deliberation. The Jordanian government, mainly represented by JAEC, suppresses realist assessments of project feasibility and management by e.g. denying risks, attempting 1 3 dominance over the nuclear regulator, or firing employees pointing out mistakes. As an environmental activist said: "Anyone who finds a mistake is out." (interview C5_E_2014). The distortion of the public discourse also consists of a strategy of utilizing and enforcing an assumed lack of technical knowledge. The government presents the average Jordanian citizen's lack of technical knowledge as a given, surrounds the project with an aura of top secrecy, and exclusively refers the matter to experts. JAEC's leadership has repeatedly been accused of not taking public concerns seriously: in 2012, its director was reportedly recorded calling the opponents of the nuclear energy program "garbage men" and "donkeys" (Seeley, 2014). Public distrust grew when deadlines for project milestones were repeatedly changing and sites were shifting, leading to people starting to feel they were getting misleading information by JAEC. JAEC, in return, argues that those concerns were based on a lack of understanding of the nuclear industry (interview JG1_2014). JAEC also refers to all relevant facts being prominently laid out in its (rather optimistic) White Paper on Nuclear Energy, which the international consultancy firm Worley Parsons had written in close cooperation with JAEC, to end emerging discourses at an early stage. This is why the strategy of distorting public deliberation represents discursive power (mainly agenda setting).
The Jordanian government further distorts public deliberation through two interlinked mechanisms: co-opting stakeholders, and concentrated efforts of securing the support of domestic key constituents. According to one of our interview partners with ties to an influential tribe, the Jordanian government offers key tribal representatives positions at JAEC and project-related companies to change their critical position in the public discourse, prompting local observers to conclude that public figures were no longer representing their local communities. On a tribal assembly one of us attended, influential tribal representatives discussed the project, voicing their concerns regarding the safety of the plants. One representative alleged that some of the influential tribes had been "bought-off" by the Jordanian government, and had seized representing the interests and arguments of their constituents in the ongoing discourse (PO1_2014). At the same time, the Jordanian government had shifted the NPP site to an area inhibited by a less influential tribe, which can be seen as a classical instrument of authoritarian politics. To accommodate another influential societal interest, the government-according to some of the project's opponents we spoke to-has instructed imams to preach on the benefits of nuclear energy (C5_E_2014). This is to add some religious legitimation to the otherwise rather technical and economic arguments of the government. These mechanisms clearly represent an expression of discursive power and materializes across all three faces of power according to Lukes (2005).
3. Delegitimizing opponents and critics. Closely linked to the Jordanian government's strategy of distortion of the public discourse are its mechanisms geared towards delegitimizing the project's opponents and critics. This delegitimization happens through undermining the opponents' credibility through presenting them as acting against the national interest, or their portrayal as destroying the Kingdom's "golden opportunity" to enter the global nuclear hub (interviews C1_E_2014, C2_E_2014). These efforts are complemented by exploiting popular images of the enemy, as criticism is framed as a wider conspiracy to undermine the nuclear program, and as such an effort to serve Israel's interest (Luck, 2012). This was also reported by an interview partner, who explained that both the King and JAEC's director Toukan had repeatedly and more or less directly said that those opposing the nuclear program were directed by Israel in terms of they were serving the Israelis by opposing the nuclear program (interview C2_E_2014) In the Jordanian context, this claim represents a very strong form of discursive power (mainly agenda-setting power), as it is suited to discredit anyone participating in any public discourse.
4. Connecting to grievances. The Jordanian government also uses a discursive strategy that presents nuclear energy as the only solution to social grievances and being without alternatives. This strategy is an expression of both agenda-setting and, to a certain extent, ideological power and consists of exaggerating the economic benefits of nuclear energy whilst promising economic development. The economic benefits linked to the project are e.g. presented as leading to reducing the budgetary burden of energy imports, and even rendering Jordan an energy exporter (Petra, 2014). Furthermore, JAEC frames nuclear energy as a precondition for economic and social development, promising it will ultimately lead to a better life of ordinary Jordanians. Accordingly, JAEC has formulated its mission to: contribute to Jordan's economic and social development through the use of nuclear energy in the elec-tricity generation, desalination, and other peaceful purposes (Jordan Atomic Energy Commission, 2019).
This can be seen as directly related to the events of the "Arab Uprisings", which saw a series of popular demands partly related to socio-economic grievances (see, e.g., Beck & Hüser, 2012). 5. Creating time pressure. By ruling out alternatives and emphasizing energy insecurity, the Jordanian government further hinders open public deliberation on the project. Again, this strategy to exercise discursive power is mainly an expression of agenda-setting power. Given the urgency attached to solving Jordan's (undisputed) energy crisis, the government repeatedly stressed that the project would have to proceed in light of the difficult situation the Kingdom was going through. The nuclear energy project is commonly presented as a national priority and a developmental state project. In all our interviews with governmental officials, it was presented as the only solution to Jordan's energy insecurity, which had to be tackled quickly. One interviewee repeatedly referred to Jordan's "horrendous energy demand and energy dependency. We import a lot of energy." (interview JG_3_2014). A member of the Jordanian Senate-senators are appointed by the King-explained: … it is clear that action was to be taken to address the energy challenges in the country. The Government has decided that nuclear energy is the only way to meet the energy, water and economic challenges […] in addition to bridging the gap of water needs for Jordan which is increasing dramatically in the light of the inflow of Syrian refugees (written statement JG_4_2016).
Whilst activists opposing the nuclear energy project argue that given Jordan's lack of water resources speaks against the construction of nuclear power plants, which would be located in the desert and require a large amount of water for Fig. 2 (continued) cooling, JAEC officials use the lack of water resources as an argument in favor of nuclear energy: …Jordan is considered one of the five poorest countries in the world […]. With rising population numbers the volume of water deficit will increase to more than 500 million cubic meters annually […]. This will cause acute shortages disabling if not devastating the agricultural and economic development, as well as posing significant risks to the lives and welfare of the citizens. (Jordan Atomic Energy Commission, 2011).
As this quote illustrates, nuclear energy is not only presented as a solution to a problem, but as a solution to a serious matter of particular urgency, which contributes to constraining potentially time-consuming public deliberations on the suitability of nuclear energy given the Jordanian context.

Strategic Mirroring as a Central Strategy of the Russian Government
Next to the Jordanian government, the Russian government dominates the scene and exercises discursive power by using two main strategies: political flanking and offering support beyond the project (see Fig. 3), which form its dominant strategy of strategic mirroring of the Jordanian government's position in the Jordanian discourse(s). In close interaction with the state enterprise Rosatom, the Russian government picks up central themes within the Jordanian discourse(s), focusing mainly on the narrative of the Jordanian government. Especially in the earlier project stages, the Russian government assured its Jordanian counterpart that it would not have to worry about funding, or guarantee its share financially. On a high-level meeting of the Jordanian King and Sergei Kiriyenko, head of Rosatom, JAEC chairman Toukan summarized the agreement with the Russian partner: The government will not be required to provide any financial guarantees for funding, which will be extended from abroad… the government will also not take any risk in drawing up measures for financial guarantees (The Jordan Times, 2015).
It proactively addresses this major concern of the Jordanian partner, who struggles with a continuously worsening budgetary deficit. On top of that, it proposes media cooperation to foster the public acceptance of nuclear energy in Jordan. Governmental officials also repeatedly take up arguments prevalent in Jordanian media. They refer to the potential of Jordan benefitting from Russia's decades of nuclear expertise, or repeat the Jordanian government's promises. During intergovernmental meetings, Russian government officials commonly highlight Jordan's role in the region, and acknowledge the impact of the high number of Syrian refugees Jordan is hosting.
Rosatom can hardly be distinguished as an independent actor in this setting. As a state enterprise, it interacts closely with the Russian government whilst officially highlighting its distance to politics (Foy, 2017). Its main strategy, too, is the strategic mirroring of the Jordanian government's position in the Jordanian discourse(s).
Rosatom has correspondingly been widely ignorant of local communities. The great majority of our interview partners stated that Rosatom had remained completely inactive and silent in terms of stakeholder involvement. Stating that Rosatom was involved in the project for political reasons, a tribal leader referred to the Russian project partner saying: There were no consultations with us beforehand, the tribes knew about it [the project] through the media. (interview C3_T_2014) Figure 4 provides the data structure for Rosatom. In line with the governmental position, Russia's quasi-Ministry of Energy Rosatom addresses Jordan's concerns for finance and designs its offers particularly flexibly. More specifically, it takes up JAEC's Build-Own-Operate equation and contributes a substantial amount of funding-49.9%-to the costly project. The project, which costs around of quarter of Jordan's GDP ($ 10 bn), otherwise would long have been stopped simply for lack of funding. Rosatom designs its offer flexibly by offering assistance in the provision of intergovernmental loans beyond the initial funding agreement, and, at least according to the Jordanian side, expecting a rather modest profit margin in the related Purchase-Power-Agreement. In moving beyond project basics, and thus addressing Jordanian technical and operative concerns, it provides scholarships to Jordan's nuclear physicists or invites Jordanian managers to participate in international training courses for nuclear newcomers (Rosatom Communications Department, 2016). The emphasis on knowledge transfer mirrors the Jordanian government's dominant understanding of project partners' responsibility in this project, which a Jordanian senator we interviewed summarizes as local capacity building: I strongly believe that multinational corporations' main responsibility is to build local capacities when implementing mega projects (written statement JG_4_2016).
At a meeting with the King, and in line with several other public statements made, Rosatom CEO Kiriyenko refers to the Jordanian partners' demands for knowledge transfer quite explicitly, and.
stressed his country's keenness to increase the number of Jordanians granted scholarships to pursue their education in nuclear energy-related fields in Russia to 1 3 30 each year, according to the statement (The Jordan Times, 2015).
Aware of international concerns for nuclear waste management, it also offers Jordan a rare option of returning nuclear fuel waste to Russia. Furthermore, its CEO Kiriyenko promises that Rosatom will exert all efforts to attract international investors, which is a particularly valuable project component given Jordan's budgetary concerns. Several interview partners also highlighted the issue of "under the table money" (e.g. interviews C1_LC_2014, C2_E_2014). Prominent figures within Jordan's anti-nuclear movement thus argue that it was corruption that caused the Jordanian government to ignore other energy options, as the "nuclear lobby" kept influencing JAEC: The whole program is built on corruption and wasta 1 […] Where did all the money go?[…] What is happening now in Jordan can serve as an international model of how a nuclear lobby can rape a country (interview C5_E_2014).
On the other hand, there seems to be continuous pressure on JAEC to accelerate the project: The Russian government really, really wants the project […] They really pressure the Jordanian part to accelerate the project […] The Russian government wants to invest more in the region to build political relationships […] They have a very tight schedule to manage the project. So they do not care to communicate with the public at all (interview C6_ET_2016). Rosatom's Director General and other high-ranking executives have engaged in intensive diplomacy, e.g. when addressing international fora, such as the IAEA, on behalf of the Russian government (see, e.g., Kirienko, 2011). Historically, Russia had tied its nuclear exports to (geo-) politics, e.g. when publicly threatening Ukraine "with a cut-off of Russian nuclear fuel if it continued down the path toward an EU Association Agreement" (Vlcek & Jirusek, 2015: 914). Russia is interested in pushing its domestic nuclear industry with the help of an export-oriented business model that particularly appeals to Middle Eastern governments. Russia profits economically from revenues out of the PPAs and "the use of a predominantly Russian supply chain for construction" (Cottee & Elbahtimy, 2016). On a geo-political level, Russia, through the complexity of associated project management, has the opportunity to establish "long-term economic links with Middle Eastern powers", a goal Russia defines as strategic (Cottee & Elbahtimy, 2016). The geopolitical relevance of the project to the Russian government, as well as Russia being an autocracy itself, may explain why it presents itself as the second most dominant actor (see Fig. 1), and why it widely evades public deliberation with civil society.
Accordingly, the Russian government and Rosatom strategically mirror the Jordanian government's position which holds that public acceptance of a nuclear energy project depended mainly on the "right" kind of information. This behaviour represents an expression of discursive power, mainly in the form of agenda-setting power, as it almost exclusively refers the matter to "experts". In an interview with BELTA, the Belarussian news agency, Kirill Komarov, Deputy Director General of Rosatom explained the rationale, referring to the importance of public acceptance for the development of nuclear energy: As the main problems are only due to lack of information, a nuclear power plant itself has nothing dangerous, the hazard is mainly in the heads of the people who do not simply have sufficient information. When people are given such information, there are normally no problems (BELTA, 2013).
It is important to note that this understanding of public deliberation is shared by central actors of the Jordanian government. As we have shown above, the Jordanian government contributes to the distortion of public deliberation by framing the topic as an exclusive matter of experts, whilst utilizing or even enforcing a lack of objective knowledge. Correspondingly, on a visit to Russian NPP in Leningrad organized by Rosatom, the permanent representative of Jordan to International Organizations in Vienna said: It is my third trip of five organized by Rosatom for ambassadors in Vienna….More and more we make sure that nuclear technologies are valuable […]. We can supply our population not only with information which confirms attractiveness of nuclear power but speak about the fact that involvement in the use of these technologies may bring benefits for environment (Rosatom, 2017).
As both the Jordanian government and the Russian government and Rosatom seem to share this understanding of public discourse, which foresees a rather passive, information-receiving role for the public, they jointly abstain from involving the latter in the sense of Political CSR.

Do in Rome…: Evading Public Deliberation as Central Strategy of the Korean Government
Figures 5 and 6 provide the data structures for the Korean government and Daewoo E&C respectively. The Korean government employs one major strategy: evading public deliberation. This behavior represents a conscious decision by the Korean government to adapt to local circumstances and thereby is an expression of discursive power, again particularly in the form of agenda-setting power. In order to create favorable conditions for project implementation, the Korean government is mainly concerned with supporting the Korean consortium by providing competitive funding, particularly through soft loans. This resembles the Russian government's strategy of promoting exports. Another similarity we observe is the Korean government leveraging instruments of its development agency KOICA, e.g. in ). Yet through actively avoiding public engagement with regard to the nuclear energy project and referring to the particularities of the host country context, the Korean government, although democratic, sticks to its strategy of evading any participation in any public discourses surrounding the nuclear energy project. Korean diplomacy builds on a notion of public diplomacy, meaning that the way Korean governmental agencies engage with foreign populations depends on the host country's strength of government and the corresponding strength of the public's voice. This leads to a strictly relativist adaptation to context, i.e. involvement of the public in democracies, and an evasion of engagement with the public in non-democratic contexts. The Korean diplomat interviewed for this article accordingly emphasizes that in his private opinion, the Korean government expects companies to behave according to the national context. They need to act according to local conditions, local moods (interview KOR GOV_D_2016).
Export contracts in this specific industry generally receive a lot of governmental support, e.g. through loans, formal state visits, guarantees or other types of diplomatic action (Vlcek & Jirusek, 2015: 915).
In Korea, nuclear energy is described as a strategic priority both domestically-25 reactors provide around a third of the country's electricity-and internationally, with the South Korean government aiming at becoming the world's third largest supplier of nuclear technology (World Nuclear Association, 2021). Currently lacking significant military or geopolitical interests in the region, it frames the region's significance as a pure matter of trade and investment (Azad, 2013). Accordingly, the country has increasingly pursued new export opportunities for its nuclear expertise. The export of nuclear energy facilities, technology and know-how is a central pillar of its national economic growth strategy (Kosch O'Donnel, 2013). In this context of a non-democratic host country and a nuclear export-driven home country with a pronounced tendency to adapt to host country contexts, public deliberation with all stakeholders affected is strategically evaded.
Another publicly silent, yet central, actor is the Korean stock-listed firm Daewoo E&C, the private company within the Korean consortium. Its major actions can be described as approaching the structures encountered opportunistically by avoiding public engagement and remaining under the radar, and thus joining the Korean government in evading public deliberation. This behavior represents a conscious decision by Daewoo E&C is an expression of discursive power. This is also in line with the Korean government's approach to public diplomacy, with distinguishes between democratic and non-democratic host country contexts. Daewoo remains as silent as possible. Local communities were not informed prior to the beginning of construction works on the JRTR. Furthermore, as the overview over this article's data sources illustrates, there are few public references to Daewoo's project involvement, and local activists conclude: "We don't know what they are doing, they are just invisible" (interview C1_LC_2014).
Whilst DAEWOO E&C highlights having won the contract for the JRTR construction as a historical success (DAEWOO E&C, 2016b), locally, it remains completely invisible in terms of stakeholder involvement. At one point, this triggered drastic reactions of the local communities: when construction on the JRTR began without having informed residents, members of the local communities burned down the offices of the South Koreans and started a series of demonstrations (interview C1_LC_2014): 150,000 people live in the area of the research reactor, without even knowing […] They [referring to the government and DAEWOO E&C] use the people´s lack of knowledge […] When our people were told what was happening at JUST, they went to the JRTR construction site, and they were so outraged seeing the works had already started.
The consortium's members can be understood as selfinterested agents of a government complying with or benefitting from a national economic interest: expanding South Korea's nuclear export industry, whilst approaching political structures opportunistically.

Summary: Strategies Countering Public Demands of Civil Society for Deliberation
We have shown that the most dominant actor, the Jordanian government, does not exclusively draw on coercive power to repress public deliberation, as one might have assumed. It also applies various faces of discursive power, which mainly materializes as agenda-setting power. It takes active measures to distort public deliberation through ensuring the dominance of the governmental narrative, utilizing and enforcing a lack of knowledge, co-opting stakeholders and ensuring the support of key constituents rather than engaging in meaningful dialogue with all stakeholders. Furthermore, it hinders genuine, open-ended public deliberation through the delegitimization of opponents and critics of the nuclear energy project, e.g. by exploiting widespread images of the enemy. In parallel, the government connects the project to social grievances, raising the stakes for critics to object to the project as it has repeatedly been framed as the only solution to Jordan's serious problems. This might indicate a discursive contribution to its ideological power, as stakeholders might thus be influenced to take a position contrary to their actual interests. However, given our position as researchers outside of the Jordanian context, we refrain from such a judgment, as it would include evaluations and attributions we are hardly competent to make.
Finally, the Jordanian government discursively creates additional time pressure further constraining the discourse, hinting at the urgent need to address Jordan's energy insecurity as quickly as possible. The authoritarian government of Russia, together with state-owned Rosatom, avoids engaging in public deliberation with the Jordanian stakeholders. It shares this strategy with the democratic government of Korea and the Korean consortium and Daewoo E&C. Whilst the Korean partners' central strategy can be summarized as evading public deliberation, and the corresponding adaptation to the non-democratic context, the Russian government primarily engages in a rather strategic mirroring of the Jordanian government's position in the Jordanian discourse(s). Overall, these strategies mostly consolidate the agendasetting power of the governments involved in the project.

Reflections on the Boundary Conditions of Political CSR in Non-Demoratic Contexts
In the following, we critically reflect the empirical findings of our study in the light of central contextual assumptions of Political CSR. We seek to contribute to a more contextsensitive development of Political CSR, as we discuss the implications of non-democratic political-institutional contexts and their distinct power asymmetries for public deliberation. We find that the various actors' interactions with the political-institutional macro-context lead to a situation where the most central provision of Political CSR-corporate engagement in public deliberation to ensure moral legitimacy-faces close to insurmountable constraints, and is not in the direct interest of the MNCs involved.
MNCs are thus commonly described as particularly powerful actors (e.g., Ruggie, 2018), and they are accordingly attributed a greater social responsibility than, for instance, small and medium-sized enterprises that also engage in international business. Our study, however, indicates that even MNCs-presupposed they sought to engage in responsible business behaviour in the sense of Political CSR-would face considerable contextual challenges if they wanted to engage in public deliberation. In our case study, however, the MNCs involved did not attempt to obtain moral legitimacy (Suchman, 1995;Palazzo & Scherer, 2006) in the eyes of the civil society affected by the project, and sought to evade public deliberation altogether. As we have shown, the MNCs and the foreign governments have either evaded an engagement in public deliberation due to a relativist strategy of adaptation to context (Korea), or through strategically mirroring the Jordanian government's position in the discourse surrounding the nuclear energy project (Russia). We find that this evasion is not mainly a result of the nondemocratic host country context. It also results from their self-interest and strategic priorities, which Political CSR research has only begun to take into account.
Political CSR, public deliberation and power in non-democratic contexts. The central idea of Habermasian deliberative democracy Political CSR normatively builds upon is that corporations ought to "engage in extensive high-quality communication in order to gather as many different perspectives as possible, so as to make the most inclusive, rational, and respectful decisions possible" (Sabadoz & Singer, 2017: 188). For political decisions to be legitimate, they need to be connected to open-ended exchanges of arguments that take place within high-quality processes of deliberation (Curato et al., 2017). The emphasis lies on mutual justification and the responsiveness of political decision-making bodies to such discursive exchanges of perspectives. A central underlying idea of deliberative democracy is that the collective communicative power exercised through deliberation helps "neutralizing" coercive forms of power (Habermas, 1977) and puts "collective decisions on a footing of common reason" (Cohen & Rogers, 2003: 242). Deliberative democracy scholars have long urged to take marginalized actors and power asymmetries in deliberative processes into account (e.g., Young, 1996).
Citizens of democracies experience various forms of power asymmetries, too (Button & Mattson, 1999). Correspondingly, power imbalances hindering inclusive stakeholder deliberation have increasingly been addressed in the Political CSR literature (see, e.g., Moog et al., 2015;Sabadoz & Singer, 2017). In non-democratic contexts, however, these power imbalances are decisively more pronounced, as political institutions are by default designed rather exclusively, and civil society demands for public deliberation are in most cases met with pressure, if not violent repression (Svolik, 2012;Sassoon, 2016).
This adds another nuance to the ongoing debates on the role of governments in Political CSR: governments often remain center-stage in democracies as well (see, e.g., Moon et al., 2010;Gond et al., 2011;Knudsen & Brown, 2015). Governments' political agenda quite visibly reaches into the rationale of corporate actors-especially when state-owned corporations are concerned (see, e.g., Schrempf-Stirling, 2018). Extending the literature, we conclude that not only do governments persist, e.g. in deliberately leaving central issues of public concern unregulated, but they also use sophisticated discursive strategies to retain their positions of power. In non-democracies, we need to acknowledge that the repression of civil society demands for public deliberation is, by default, a matter of political choice and a characteristic of the political system. In that context, MNCs can often be assumed to approach these structures opportunistically, rather than even attempting to mitigate the particularities of authoritarian decision-making, e.g. the repression of civil society demands for public deliberation. What this implies, then, is that we need to overcome the dominant reading of Political CSR activities of MNCs being linked to varying degrees of state capacity. Rather than basing further analyses on often implicit assumptions on weak or limited statehood, a closer look at the concrete design of political-institutional contexts is warranted. This would allow for more nuanced analyses that provide the basis for both context-sensitive theoretical reflections and managerial recommendations. Finally, this would further scholarly attempts at a more empirically grounded appreciation of the limitations and potential of Political CSR in contexts that typically lie within the shadows of CSR research.
Contrary to our initial expectation that violent repression of public deliberation as an expression of coercive power would be the most commonly employed strategy in this nondemocratic context, we found that the most salient form of power was discursive power that mainly materialized as agenda-setting power. The distortion of public deliberation in terms of an authoritarian agenda-setting was the most dominant theme. For example, the government (allegedly) engaged traditional authorities that hold their own, limited power such as Imans or tribal representatives to influence the public discourse on the nuclear energy project in their favor. In our findings, we refer to that strategy as a strategy of co-optation that is quite common to non-democratic political systems. Non-democratic governments thus use their agenda-setting power beyond their decision-making power (understood basically as the power of governments to legitimately make decisions on behalf of the people) (Lukes, 2005) and employ sophisticated, discursive strategies that lead to an exclusion of several voices from public deliberation. By dismissing opponents of the project as incompetent and lacking technical expertise, the Jordanian government also exercises its discursive power over and in communication (Pellizzoni, 2001) and does not include different perspectives even on strictly technical issues. Such exclusion of perspectives can also occur in democracies-however, it would be far less systematic. Furthermore, institutional remedies and safeguards, such as plan approval procedures or legal proceedings, would be available to citizens. Therefore, even if public deliberation occurs in nondemocracies, it will necessarily lack central preconditions for legitimacy, e.g. the inclusion of different perspectives into such deliberative exchanges. In political-institutional systems that fundamentally rely on the exclusion of the majority of the population from political decisions and the dominance of the governmental interpretation of issues, any deliberation with and among the public quickly becomes a threat to authoritarian rule, regardless of its specific content (Stockmann et al., 2020). We find these more subtle forms of discursive power to be more prevalent and efficacious in our empirical setting than purely coercive power through e.g. violent repression. This echoes studies on deliberative democracy in practice that problematize power resulting from inequalities amongst participants (Hendriks, 2009: 181).
Political CSR and normative implications of the politicalinstitutional context. Even if MNCs were genuinely interested in engaging in public deliberation with civil society in non-democracies, they might not be able to in the way the literature on Political CSR in democratic contexts suggests. Therefore, simply applying "Western" conceptualizations of discursive formats like a Multi-Stakeholder Initiative (MSI) to a non-democratic context would also be futile, as an MSI would already lack legitimacy due to its inevitable lack of input legitimacy, e.g. with regard to the criteria of inclusiveness (Mena & Palazzo, 2012). However, rather than resigning given these grim prospects for deliberation, a first step forward could be to account for political context by being "less rigid, demanding, or expectant with regards to how participants ought to communicate and deliberate" (Sabadoz & Singer, 2017, p. 201).
Non-contextualized Political CSR has limited authority outside of its standard cases, and especially in non-democratic contexts that are particularly hostile towards public deliberation and inclusive, non-discriminatory public good provision. Ehrnström-Fuentes' suggestion of a notion of pluriversal legitimacies which "acknowledges the existence of different place-based social imaginaries that structure not only the perception of what is deemed morally right in a particular community but also the possibilities for different stakeholders to engage in meaningful dialogues" (Ehrnström-Fuentes, 2016: 458) is one we share, but not without reservations. Rather, we propose taking a middle ground between relativism and universalism by acknowledging that context matters and sets boundaries for concepts and normative prescriptions, whilst upholding principles we also deem appropriate for democracy-based stakeholders even when facing challenging political-institutional contexts. As a guiding, normative principle, public deliberation remains a normative provision worth exploring in different contexts. Just because a government demonstrates great repressive capacity, thus marginalizing the voices of civil society, we should not assume that there are no civil society demands for public deliberation. Furthermore, as the "Arab Uprisings" have shown, non-democracies are not exempted from societal and political, sometimes disruptive, change (see, e.g., Gerges, 2015). Given the public resistance of a multitude of Jordanian civil society actors-ranging from environmentalists to tribal representatives-against the nuclear energy project, one should be careful not to rule out the normative appeal of Political CSR by confusing the repression of public deliberation with a popular lack of interest in it. To the contrary, basic democratic ideas can be found across different cultures and political systems (Sawani, 2014: 351), and corporations should find ways to address them adequately.
Other pressures, e.g. competition for governmental orders, and the limited prospects of diffusion of values and principles of Political CSR into the host country context notwithstanding, MNCs should at least amplify the voices of stakeholders wherever possible, and actively look for spaces to include different perspectives and voices. Therefore, we propose to initially focus on advancing more technical, pragmatic discourses in this context, as these are primarily concerned with the effectiveness of means, rather than the goodness of ends (ethical discourses) or the generalizability of norms (moral discourses) (Gilbert & Behnam, 2009, p. 224). Another step forward can be taken by engaging with existent, and more traditional, forums of public deliberation such as the tribal assembly we attended as part of our data collection, or to emphasize the importance of stakeholder involvement in high-profile intergovernmental talks that accompany such state projects.
We expect this challenge to be considerably greater in fully consolidated and much more repressive autocracies that are rich in natural resources (e.g., Saudi Arabia or the United Arab Emirates) than in less repressive non-democracies like Jordan, which also heavily depends on foreign funding. Whilst MNCs should take managerial implications of nondemocratic contexts for their responsible business conduct seriously, our concepts of Political CSR need to develop further contextual sensitivity. Going beyond merely recognizing its limitations in challenging contexts, this can translate into designing contextually adapted criteria and formats for public deliberation that are, as written above, less rigid and formalized than in ideal theory. Further research should go beyond identifying boundary conditions, and develop concrete and practical managerial recommendations for these particularly challenging contexts demanding the most of corporations' Political CSR.

Conclusion
We contribute to a more nuanced theoretical understanding of how power, as it is embodied in the non-democratic political system forming the context of this case, materializes in (discursive) struggles surrounding the project (see, e.g., Banerjee, 2010). Our study shows that stakeholders in a non-democratic context do not exclusively draw on coercive power to repress public deliberation, but also apply various forms of discursive power reach their strategic objectives. Utilizing rich data on an extreme case of governments and MNCs shaping a context particularly inhospitable to responsible business conduct in a non-democracy, we make three contributions to the theoretical advancement of Political CSR. First, we provide unique and rich data on an underresearched non-democratic political-institutional context to address calls for empirical insights suitable to fill a gap identified in Political CSR's expanded research agenda (Scherer et al., 2016). Second, and having deconstructed central assumptions of Political CSR research on political context, we contribute to further developing Political CSR into a more context-and power-sensitive concept. As we have shown, sensitivity to context and a clear distinction between different forms of coercive and discursive power are key, especially when studying phenomena linked to critical issues that can be subsumed under the label of socially irresponsible business conduct (Linstead et al., 2014: 178). Third, we contribute to the growing literature on the role of governments in the governance of business conduct (Kourula et al., 2019), whose insights are of paramount importance for the further development of Political CSR. Based on our findings, we argue for a more context-sensitive discussion of the role of governments regarding MNC's business conduct. Further research should not only explore the role of governments, but also make explicit what type of governments are the subject of analysis, i.e. democratic or autocratic regimes. Assumptions on the political-institutional context of Political CSR, in our view, need to be critically reflected and articulated explicitly, especially since the majority of critics of Political CSR, too, base their analyses on the implicit assumption of "Western" MNCs in democracies or contexts of limited statehood.
Our study is not without limitations. Given the challenging research context and the non-responsiveness especially of corporate actors we asked for an interview, our interview database is dominated by accounts of civil society. Both the governmental and corporate perspective were difficult to account for given the challenging research context. One could argue, however, that this makes our research more sensitive to the effects of power asymmetries on marginalized groups. Furthermore, in the light of personal risks for both the researcher and interview partners, our interviews could not be recorded. Theoretically, we did not differentiate between and address the particularities of different types of non-democracies, but rather focused on some authoritarian themes. A more nuanced analysis dealing with sub-contexts could prove fruitful to determine in which challenging settings Political CSR's central pillars might unfold thus far unexpected potential. Given the "Political" in Political CSR, political-institutional contexts and their distinct configurations of power deserve closer attention.