Corporate social responsibility decisions in apparel supply chains: The role of negative emotions in Bangladesh and Pakistan

This article integrates the global value chain literature with the micro organization literature on negative emotions to explore the drivers of fear and anger among supplier factory senior managers in apparel supply chains after Rana Plaza — a major industrial disaster — and their influence on decisions on CSR practices. Based on a comparative study around Dhaka and Lahore — two key apparel manufacturing hubs — this study elucidates that supplier factory senior managers experienced similar market tensions but different social tensions after the Rana Plaza incident. Crucially, similar market tensions helped create market fear and anger, but different social tensions led to social fear and anger in Bangladesh but not in Pakistan, therefore influencing the way supplier factory senior managers take decisions regarding CSR practices. By conceptualizing communal alignment and competitive CSR, this research finally advances the global value chain literature and contributes to the current conversations on negative emotions in organizations.


| INTRODUCTION
Global value chain (hereinafter: GVC) scholars have long studied the implementation of social upgrades to improve labor conditions for supplier factories (hereinafter: suppliers) in developing countries (Barrientos et al., 2011;Humphrey & Schmitz, 2002;Mzembe et al., 2016;Ponte & Sturgeon, 2014;Welford & Frost, 2006). For apparel supply chains, the GVC literature has emphasized the need to complement coerciveness and compliance with corporate social responsibility (hereafter: CSR) practices (Jiang, 2009;Lund-Thomsen & Lindgreen, 2014;Soundararajan & Brown, 2016). These are defined as voluntary practices holding the promise to provide more rights, entitlements, and better welfare to workers (Alexander, 2020;Foerstl et al., 2015;White et al., 2017). More significantly, GVC scholars have recently acknowledged the importance of supplier senior managers.
As individuals who hold a top and representative position in their factory, they are germane because they often decide on CSR practices, collectively influencing labor conditions upstream in apparel supply chains (Fontana & Egels-Zandén, 2019;Huq & Stevenson, 2020;Klassen & Vereecke, 2012).
Nonetheless, evidence in the GVC literature of the influences on the CSR decisions of supplier senior manager in upstream apparel supply chains remains embryonic. Yawar and Seuring (2017) for instance illustrated that the lion's share of GVC studies in the Global South have taken a Western and an international buyer firm (hereinafter: buyer) perspective. This has led to little awareness of the challenges and emotional circumstances experienced by supplier senior managers (Khan et al., 2020;Lund-Thomsen, 2020). This gap is particularly problematic to understand CSR decisions because supplier senior managers, as humans, take choices that do not always follow rational justifications. They are likely to mirror inner feelings whose influence is often unaccounted for in the literature (Baden et al., 2009;Harwood & Humby, 2008;Sajjad et al., 2020;van Hoek et al., 2002).
This study argues that supplier senior managers in upstream apparel supply chains might be susceptible to negative emotions when taking CSR decisions, especially fear and anger. These have often been operationalized as a unitary construct in the micro organization literature on negative emotions not only because of their similar ability to steer the decisions of senior managers regarding organizational practices (Barnard, 2008), but also because they represent the most prevalent discreet negative emotions in organizations (Coget et al., 2011;Lebel, 2017;Lerner & Tiedens, 2006). As fear and anger often emerge because of disruptive events (Friedrich & Wüstenhagen, 2017;Hine, 2004;Huy, 2002;Shepherd & Williams, 2014), this study focuses on the ongoing tensions after the Rana Plaza incident. Since this was the largest industrial disaster of the past decade, killing more than 1100 apparel workers (Chowdhury, 2017), it is an essential consideration in studying fear and anger in upstream apparel supply chains as a whole. Rana Plaza had a major influence on the entire apparel industry in improving labor conditions (Comyns & Franklin-Johnson, 2018;Donaghey & Reinecke, 2018), but also raised new dilemmas at the level of supplier senior managers about CSR decisions (Bux et al., 2020).
This study responds to the growing calls for more evidence on supplier senior manager decisions on CSR practices in the GVC literature (Khan et al., 2020;Lund-Thomsen, 2020;Yawar & Seuring, 2017) by answering the following two questions: 1. What are the drivers of supplier senior managers' fear and anger in the apparel supply chains of developing countries after the Rana Plaza incident? 2. How do supplier senior managers' fear and anger influence their decisions on CSR practices after the Rana Plaza incident?
By integrating the GVC literature with the micro organization literature on negative emotions, this study proposes a comparative and qualitative design that is key to study human judgment and organizational practices (Dobers, 2009;Lehnert et al., 2016).
It thus draws evidence from fieldwork and 32 interviews conducted between 2016 and 2019 with supplier senior managers operating in Bangladeshi and Pakistani apparel supply chains. It specifically follows the arguments of Quazi and O'Brien (2000) and Williams and Aguilera (2008) to benchmark market and social elements when designing comparative research that involves CSR in multiple countries because they help explain managerial decisions on CSR practices.
Hence, comparing Bangladesh and Pakistan is appropriate because they used to be part of the same country and share similar social and market structures (Jamali et al., 2009;Khan, 2014;Khan et al., 2021;Mishra, 2012;Naeem & Welford, 2009;Sobhani et al., 2009).
This study contributes to the literature in three ways. First, it contributes to the GVC literature by demonstrating that social upgrades in upstream apparel supply chains can be better achieved by complementing the erstwhile GVC governance models with a more thorough scrutiny of the negative emotions of supplier senior managers, which can explain their decisions on CSR practices-from micro to organizational levels. Conceptualizing communal alignment in Bangladesh and competitive CSR in Pakistan demonstrates that these decisions hinge on the tensions that drive the negative emotional investment of supplier senior managers.
Second, the study contributes to the current debate on negative emotions in the micro organization literature by highlighting the importance of context in fostering fear and anger. The Rana Plaza incident, analyzed through the experience of supplier senior managers via a comparative study, gives insight into not only the significance of disruptive events as a source of fear and anger, but also the importance of context in filtering tensions at the market and social levels.
Finally, this study illuminates the empirical context of the Bangladeshi and Pakistani apparel supply chains, focusing on worker unrest as a double-edged sword regarding CSR practices and improving worker conditions. 2 | THEORETICAL BACKGROUND 2.1 | Social upgrades in the supplier factories of developing countries in the GVC literature GVC scholars acknowledge that economic upgrades in upstream supply chains have largely failed to translate into social upgrades that improve labor conditions (Humphrey & Schmitz, 2002;Ma et al., 2016;Mzembe et al., 2016;Ponte & Sturgeon, 2014;Welford & Frost, 2006). Many economic geography and political economy studies have highlighted the horrific labor conditions at supplier factories in the apparel supply chains (Huq et al., 2016;Huq & Stevenson, 2020), including rising inequality affecting workers (Klassen & Vereecke, 2012;Lee et al., 2019;Soundararajan & Brown, 2016). A key example is the Rana Plaza factory collapse in April 2013 in Bangladesh due to poor infrastructure, which killed more than 1100 workers, and shocked the apparel world (Chowdhury, 2017;Donaghey & Reinecke, 2018;Fontana & Egels-Zandén, 2019).
GVC scholars turned their attention to the management of labor conditions (Barrientos et al., 2011;Locke & Romis, 2010;Lund-Thomsen & Nadvi, 2010), thereby introducing three main governance models driven by buyers to realize social upgrades at supplier factories. The development of these models is of key importance because it signals a shift from coercive to more flexible and voluntary practices, especially in upstream apparel supply chains (Alexander, 2020;Foerstl et al., 2015).
These social upgrades have been first studied using a "compliance" model, which is also recognized as the "conventional" method of transactional governance (Soundararajan & Brown, 2016, p. 87). By featuring the enforcement of the codes of conduct of buyers, the compliance model is predominantly predicated upon the sanctioning and rewarding of suppliers by buyers for their social performance through better prices and manufacturing volumes (Huq & Stevenson, 2020). However, this model has been repeatedly critiqued because of its inherently coercive nature and mixed results (Baden et al., 2009;Brammer et al., 2011;Foerstl et al., 2015;Locke, 2013;Locke & Romis, 2010;Welford & Frost, 2006).
Given the importance of collaboration (Lozano & Barreiro-Gen, 2021;Murcia et al., 2020), a second cooperative model, dubbed the "cooperation paradigm" and rooted in the relational view (Dyer & Singh, 1998;Lee et al., 2019;Lund-Thomsen & Lindgreen, 2014), has been praised for facilitating social upgrades through greater flexibility and mutual commitment. For instance, Locke et al. (2007) documented how the cooperative model allowed for fixing social ills at supplier factories in Central America and Asia by combining monitoring efforts with flexible work schedules for workers. The cooperative model also improves the buyer-supplier relationship via enhancing reciprocal trust, closer collaboration, new production techniques, work processes, and supplier capabilities (Huq et al., 2016;Klassen & Vereecke, 2012;Lund-Thomsen & Lindgreen, 2014). Nevertheless, GVC scholars have cautioned that the cooperative model is unlikely to lead to sustained improvements in labor conditions because of the permanently asymmetric power relationship between suppliers and buyers (Anner, 2018) and the challenge buyers face "[in engaging] in close cooperation," given the high number of suppliers (Lund- Thomsen & Lindgreen, 2014, p. 17).
The GVC literature has also illustrated a third model: the "multistakeholder" model. This retains the collaborative features of the cooperative model but aims to lower the asymmetric power issue in dyadic buyer-supplier relationships by including the moderating role of third-party stakeholder organizations at the transnational level (Arora et al., 2020). These organizations facilitate social upgrades by cooperation among suppliers and buyers, thus assisting in the implementation and monitoring of transnational labor standards (Rasche, 2012). Although the multi-stakeholder model represents an evolution in the GVC literature and practice, it is unclear if it can remedy the labor conditions in developing countries (Arora et al., 2020).
As anticipated, these governance models demonstrate a shift from coercive to more flexible practices, ultimately encouraging CSR practices (Foerstl et al., 2015;White et al., 2017). Following Alexander (2020), CSR practices in the apparel supply chains are defined as voluntary practices enacted to improve labor conditions and, if needed, are supported by buyers through idea generation and training. In particular, the GVC literature has acknowledged the importance of better understanding CSR practices by scrutinizing supplier senior managers as key representatives in their factory. These are key decision-makers with regard to CSR practices and collectively shape labor conditions in upstream apparel supply chains (Fontana & Egels-Zandén, 2019;Huq & Stevenson, 2020;Klassen & Vereecke, 2012).
Despite the significance of supplier senior managers for CSR practices, the evidence of the influences on their CSR decisions remains embryonic. This is particularly overshadowed by the simplistic assumptions in the literature that CSR decisions fully depend on buyers from the West (Lund-Thomsen, 2020). This tenet is fleshed out by Yawar and Seuring (2017, pp. 636-637), who cautioned that "studies so far on the management of social issues in supply chains are typically driven and explored from a Western [buyer] perspective." Likewise, Khan et al. (2020) illustrated that supplier senior managers face challenges and emotional circumstances that steer their CSR decision and are often overlooked in the literature.
This study argues that the CSR decisions of supplier senior managers in upstream apparel supply chains might be influenced by their fear and anger, which are likely to manifest because of disruptive events. However, their decisions often rely on personal feelings, especially negative ones (Baden et al., 2009;Harwood & Humby, 2008).
This situation echoes van Hoek et al.'s (2002, p. 119) "emotional capabilities" and the "criticality of [the] human aspect" of supplier senior managers in supply chains. Despite their emotional role in steering organizational practices, supplier senior managers in apparel supply chains are often labeled as unethical and inhumane in the GVC literature, which has been predominantly occupied with identifying their violations (Bird & Soundararajan, 2018;Jiang, 2009;Soundararajan et al., 2018). Such generalizations demonstrate the existence of a fundamental problem in understanding the diffusion of CSR practices for workers and call for more and systematic evidence of the human and moral capital of supplier senior managers and their influence (Sajjad et al., 2020). The next section describes fear and anger as key negative emotions and their association with radical or disruptive events.

| Fear and anger as negative emotions in organizations
Negative emotions, first conceptualized in sociology and social psychology, have gained relevance in the micro organization literature as states of unconscious and adverse feelings that are shared within circumscribed groups or communities and represent a fundamental component of collective action (Elfenbein, 2007;Hochschild, 1979;Jasper, 2011). Fear and anger are the most prevalent discreet negative emotions that shape individual behaviors in organizations and are often observed in the literature as a consequence of each other to the extent of being operationalized a unitary construct (Lebel, 2017;Lerner & Tiedens, 2006;Onkila, 2015;Schieman, 2006). While they also trigger other negative emotions, such as anxiety or sadness (Huy, 2002), they hold special valence because they influence the intuitive decisions of senior managers and their organizational practices. For example, Barnard (2008, p. 425) highlighted that fear and anger are among the most important reasons senior managers shift priorities; moreover, "like a fearful CEO, an angry CEO may generate negative consequences for a company." However, the evidence on the idiosyncratic outcomes of fear and anger in the micro organization literature remains mixed, thus simultaneously conducing to dysfunctional, unintended, but also constructive behaviors. Fear has been traditionally associated with not only withdrawal, defensive efforts (Frijda et al., 1989;Lebel, 2017), denial, and indecision (Barnard, 2008), but also ethical decisions (Kligyte et al., 2013). For instance, Kleres and Wettergren (2017) indicated that the fear of climate change drives climate activists to reframe practices that mobilize others and raise awareness. Traditionally, anger can be equated with not only poor ethical choices, the imposition of power, and control (Schieman, 2006), but also with care, social order, problem resolution, and social awareness (Geddes & Callister, 2007;Lindebaum & Gabriel, 2016;Lurie, 2004). Wijaya and Heugens's (2018) Gill and Burrow (2018) showcased that international elite chefs increase their fear of poor performance through verbal and physical anger, thus prompting fear in the community and reproducing traditional cooking practices.
Fundamentally, fear and anger are likely evoked collectively due to the repercussions of disruptive or radical events that serve as stimuli (Scherer, 2005;Schieman, 2006). These events are industrial or natural disasters, or even major restructuring projects, experienced differently in various organizational fields, communities, and under different social expectations (Hine, 2004;Shepherd & Williams, 2014).
Since these events affect the social environment, as is the incident of Rana Plaza, the fear and anger occurring due to their repercussions might violate expectations, change the perceptions of people regarding organizational practices as opportunities or threats (Jasper, 2011).
For instance, Friedrich and Wüstenhagen (2017) illustrated that senior managers in the energy industry in Germany expressed anger after the Fukushima disaster in Japan, thus contributing collectively to phasing out nuclear practices. Additionally, Huy's (2002) radical change event showed that middle managers were emotionally committed when attending to employees' fear and anger which allowed the new organizational practices to flourish. Conversely, Friedrich and Wüstenhagen (2017, p. 192) cautioned that "research on the effects of negative emotions on the decision-making process of key agents within an organization is still lacking." This situation is especially the case for supplier senior managers operating in apparel supply chains and their CSR practices, as this article illustrates.

| A tale of two fields: The Bangladeshi and Pakistani apparel supply chains
Bangladesh and Pakistan are selected for their social and market similarities. This selection follows Quazi and O'Brien's (2000) and Williams and Aguilera's (2008) arguments on benchmarking markets and social elements when designing comparative studies on CSR.
Socially, Bangladesh and Pakistan represent two post-colonial societies in the Asia-Pacific and they are characterized by "cultural homogeneity" (Khan et al., 2021, p. 405). They used to belong to the same territory as "East Pakistan" and "West Pakistan," respectively, between 1955 and the liberation war of 1971, which led to high poverty in both countries (Khan, 2014;Yunis et al., 2018). As the "two leading Muslim states of South Asia," religion dictates public life and socially accepted norms in both countries (Mishra, 2012, p. 283).
Market-wise, apparel manufacturing is their most important sector; their apparel supply chains are respectively ranked second and eighth worldwide regarding exports. Due to the Multi-Fiber Arrangement, apparel manufacturing played a catalytic role in their economic development since the 1970s (Munir et al., 2017;Sobhani et al., 2009)

| Theoretical sampling
This study addresses the research questions via an exploratory and comparative study. Qualitative methodological approaches seek to uncover the social construction of respondents' meaning (Dobers, 2009;Eisenhardt et al., 2016). They are powerful because they advance theoretical knowledge on sensitive and social issues, thereby letting the subjective and human experiences of managers emerge (Lehnert et al., 2016). While focusing on the drivers of fear and anger and their influence on CSR decisions, this research responds to the call for more comparative studies in developing countries (Jamali et al., 2009) and draws on the rhetoric of supplier senior managers from similar social and market contexts (Quazi & O'Brien, 2000). Although comparative study designs on CSR are rare due to empirical complexities, they are critical for establishing the relationship between managers and their motivation for CSR practices (Williams & Aguilera, 2008).
In total, 32 semi-structured, in-depth interviews were conducted face-to-face between August 2016 and December 2019, with 16 supplier senior managers belonging to 16 apparel manufacturing firms near Dhaka (Bangladesh) and 16 supply senior managers belonging to 10 apparel manufacturing firms near Lahore (Pakistan) (see Table 1).
Dhaka and Lahore were chosen because of the density of woven and knit manufacturing (apparel) suppliers. The interviews lasted 90 min on average in Bangladesh and 60 min on average in Pakistan. Following Barnard (2008), informal conversations before and after the interview, including lunch with the supplier senior managers, were essential because they helped the authors better understand respondents' collective feelings regarding the Rana Plaza incident. The first author visited 13 out of the 16 premises in Bangladesh, and the second and third authors visited all 10 premises in Pakistan. Participant observations, detailed fieldwork notes, and memos were later triangulated with the interview data.
This article focused on primary qualitative data through interviews, as per Scherer's (2005, p. 712) point that negative emotions following a disruptive event such as fear and anger "are experiences of mental and bodily changes" and, therefore "there is no access to them other than to ask individuals to report on the nature of their experience." Given the highly sensitive nature of this study, the respondents' job titles are disclosed, but aliases are used.
Thus, to ascertain the internal validity and rigor of the sample (Yin, 2003), the participants were selected according to the following criteria: 1. All respondents had to work for first-tier and export-focused suppliers of comparable size. Small-and medium-sized enterprises (below 250 employees) and suppliers that sold locally were excluded because of their different market conditions.
2. Since supplier senior managers hold a representative function and manage their factories, all respondents had to demonstrate consolidated management knowledge, with at least 10 years of experience in apparels (some had more than 20 years of experience).
3. Only respondents who were reachable in person were recruited to enable us to observe the fear and anger of respondents while having an open, face-to-face dialog.
4. Respondents had to be willing to let the authors spend time in their factory before or after the interview to develop an impression of the facilities and working conditions. All respondents also had to be available for the follow-up questions and informal talks after the interview.

| Data collection
The data collection process occurred between August 2016 and December 2019. The first author conducted three interviews in August 2016, while observing three factory premises near Dhaka.
These were part of a broader project on CSR and private regulation in Bangladesh after the Rana Plaza incident. During the interviews, the respondents were explicitly asked about not only their feelings after the incident, but also how they felt the incident impacted their apparel supply chain, personal work, and work-related emotions. This preliminary round of data collection was key to testing the appropriateness of the interview protocol and prepare for the other interviews. The last four interviews, conducted in 2019 in two factory premises near Lahore, were key to reaching theoretical saturation, thus informing the theory to a point at which no additional data were necessary.
As shown by the data structure (Figure 1), open coding was conducted to categorize the empirical accounts of respondents' rhetoric as first-order concepts (Gioia et al., 2013). Specifically, the authors searched for recurrent claims and observations on the fear and anger of supplier senior managers after the Rana Plaza incident, as well as their influence on CSR decisions, yielding over 50 first-order code categories concepts. Hence, to reduce complexity, the first-order concepts were grouped (Corbin & Strauss, 2008). The third section answers the second research question by indicating that similar market tensions but different social tensions contributed differently to supplier senior managers' CSR decisions: (1) communal alignment in Bangladesh and (2) competitive CSR in Pakistan.

| Fear and anger due to the market tensions after Rana Plaza
The supplier senior managers similarly communicated that the Rana Plaza incident drove buyers not only to increase their compliance requirements and promote social upgrades (Donaghey & Reinecke, 2018;Huq & Stevenson, 2020), but also to lower purchasing prices at their factories' expense. Following the second-order themes (Figure 1), the market tensions from the unethical behavior of buyers after the incident led to similar (1) market fear and (2)   Market anger involved strong feelings of animosity, associated with the sense of being tricked by buyers after the Rana Plaza incident.
Market anger appeared not only in the rhetoric of supplier senior managers and merely as perceived unfairness (Kligyte et al., 2013), but also through adjectives such as "maddening" and "irritating," which contributes further to the depiction of buyers as "exploitative" and "unethical."

Market anger was expressed by 10 of the interviewed
Bangladeshi supplier senior managers, especially when touching upon the issue of prices after the Rana Plaza incident. Shakil, for instance, shouted during the interview that "the buyer is paying me a lower price, but he is asking me to increase [worker] salaries … They are sucking our blood!" As Wasi defended: The buyers have come up with higher requirements and standards. At first, it seemed like a good intention to provide the right protection to our workers.

| Fear and anger due to the social tensions after Rana Plaza
As identified by the second-order themes (Figure 1), worker unrest emerged as the main source of social tension, not only as a risk factor for buyers (Anner, 2018). The evidence shows the existence of (1) social fear and (2)  Social fear materialized as explicit feelings of restlessness and discomfort for supplier senior managers because of their negative experience with worker unrest after the Rana Plaza incident and the concerns over more violent protests. Similar to market fear, social fear was attributed to an increasingly uncertain social context (Schieman, 2006) and surfaced in the rhetoric of supplier senior managers through adjectives such as "afraid," "frightened," and "terrified." The idea of Rana Plaza could start social problems also here in Pakistan, which is also a matter of concern for me. But as [a] human, I understand. It's tough for them.

Twelve of the supplier senior managers interviewed in
Analogously, social anger was identified during the interviews as an intense feeling of animosity associated with the belief that, after the Rana Plaza incident, unrest was fomented by unidentified third-party organizations and the media. Similar to market anger, social anger emerged in the rhetoric of supplier senior managers through adjectives such as "upsetting" and "infuriating," but also through accusations of "opportunism" toward third-party organizations and the media (Friedrich & Wüstenhagen, 2017  [business] society. Everybody is happy, and it is the best way to win in the market, without extra problems.
We are helping others. We are also helped by the other factories.
As Raju also clarified: With CSR, there will be unrest. Competitive CSR among the interviewed Pakistani supplier senior managers resulted from market fear and anger after Rana Plaza, but not from social fear and social anger. The evidence helped formally conceptualize competitive CSR in Pakistan as the implementation of CSR practices by supplier senior managers to navigate the market tensions after the Rana Plaza incident. Due to the lack of social fear and anger, CSR practices were perceived to boost competitiveness because they increased the satisfaction of workers and buyers without creating antagonism and retaliation effects with other supplier senior managers.
Hence, the supplier senior managers interpreted CSR practices as an opportunity to innovate (Coget et al., 2011;Friedrich & Wüstenhagen, 2017 Contrary to Bangladesh, two supplier senior managers in Pakistan highlighted that CSR practices could lead to minor but positive competition issues. According to Nadeem, "CSR could create competition but not unrest. Doing more CSR even after Rana Plaza will not be risky for the business community." Farooq pinpointed the following: CSR can create positive competition, and others will try to follow. They need to adapt accordingly. I will give you an example. Our recent CSR created a lot of buzz in the apparel sector, especially in Lahore and Kasur. Other workers wanted to get jobs in our factory. The neighboring factories were afraid that their competent and talented staff could move to us. So, they also increased their salaries and benefits. I think the biggest winners of CSR are the workers. CSR is a good thing. Although all supplier senior managers in Pakistan highlighted that CSR could provide some market benefits, their responses differed from the previous GVC accounts that CSR can lead to better prices (Huq et al., 2016;Ma et al., 2016). In fact, they were openly skeptical that CSR practices could directly alter buyers' prices. Zain exemplified the situation as follows: Well, buyers are quite content with CSR. If we do more than what they ask for, they are obviously happier, and F I G U R E 2 Result summary this will improve our standing in the market. But, will they give us more money only on the basis of our CSR? [laughter] I don't think so. CSR performs well if we are also good at manufacturing productivity, like [timeliness] and quality.

| DISCUSSION AND CONCLUSIONS
This study responds to recent calls to investigate the CSR decisions of supplier senior managers in upstream apparel supply chains of developing countries (Khan et al., 2020;Lund-Thomsen, 2020;Yawar & Seuring, 2017) and bridge micro and organizational dimensions by a comparative study design that stresses market and social dimensions as explanatory variables (Jamali et al., 2009;Quazi & O'Brien, 2000;Williams & Aguilera, 2008). Meanwhile, it integrates the GVC literature with the micro organization literature on negative emotions. It specifically explores the drivers of fear and anger among supplier senior managers in apparel supply chains after the Rana Plaza incident.

| Theoretical contributions
By offering insight into the fear and anger of supplier senior managers as enduring influences of their decisions on CSR practices in upstream apparel supply chains, this study provides an integrative and more comprehensive approach to the human and collective dimension of negative emotions in the GVC literature. GVC scholars have been anchored in rational and buyer-driven governance models (Humphrey & Schmitz, 2002;Ponte & Sturgeon, 2014), which have limitedly improved labor conditions (Lee et al., 2019;Mzembe et al., 2016;Welford & Frost, 2006). Conversely, rather than substitute GVC governance models, this study highlights the importance of complementing them to increase efficiency and realize social upgrades.
This study agrees with Klassen and Vereecke's (2012)  argument on the importance to detail the inner side of supplier senior managers as explanatory variables for their decisions (Baden et al., 2009;Harwood & Humby, 2008;Sajjad et al., 2020), thus serving to understand the collective state of social upgrades in apparel supply chains.
Although the need of complementing the coercive and compliance governance models has been voiced in the GVC literature (Brammer et al., 2011;Locke, 2013;Locke & Romis, 2010), this study broadens the debate on the effectiveness of CSR practices as voluntary modes of social upgrade in apparel supply chains (Alexander, 2020;Foerstl et al., 2015;White et al., 2017). Further, it cautions that CSR practices will likely not improve labor conditions without a supply chain analysis that includes the voice and perspective of supplier senior managers and goes beyond the traditionally Western and buyer-driven mode of dealing with social upgrades (Lund-Thomsen, 2020). Appraising market and social tensions in upstream apparel supply chains and detecting supplier senior managers' negative emotions at the collective level-often intensified by disruptive events such as the Rana Plaza collapse (Chowdhury, 2017;Donaghey & Reinecke, 2018)-is critical for GVC scholars to avoid inertia and even cynicism concerning CSR decisions, as well as for the supplier senior managers in Bangladesh.
Hence, this study contrasts the conventional accounts in GVC literature that assume buyers in apparel supply chains drive CSR practices, and supplier senior managers are inherently unethical and act as followers (Bird & Soundararajan, 2018;Jiang, 2009;Soundararajan et al., 2018). By conceptualizing communal alignment and competitive CSR, this study further contests these views for failing to account for the embedded and conflicting tensions at the contextual level that drive the negative emotions of supplier senior managers and differentiate their CSR decisions.
Second, this study contributes to the current debates on fear and anger in the micro organization literature on negative emotions, thereby providing evidence in support of their impact on organizational practices (Barnard, 2008). Specifically, it shows that CSR deci- Putting a spotlight on the Rana Plaza incident as a source of CSR dilemma and tension that drives fear and anger, this research follows the extant studies that highlight the importance of context in determining the extent of tensions (Geddes & Callister, 2007). Different from the previous accounts on fear and anger which focus on disruptive events within one context-affecting a specific group of decisionmakers in the same context (Friedrich & Wüstenhagen, 2017;Huy, 2002)-this study is the first to report the repercussions of a disruptive event on the organizational practices of homogeneous players from the same sector but different contexts-one where the event occurred (Bangladesh) and one experiencing only indirectly (Pakistan).
In so doing, it broadens Kleres and Wettergren's (2017) explanation of the socio-material contextualization of fear and anger by indicating that the same events are universally disruptive but locally filtered.
Through the conceptualization of communal alignment and competitive CSR in particular, this study connects with Jasper's (2011) remark that fear and anger after a disruptive event can help frame organizational practices as opportunities or threats. In the case of Bangladesh, for instance, the study corroborates the general argument that fear leads to withdrawal and defensive efforts (Frijda et al., 1989;Gill & Burrow, 2018;Lebel, 2017;Onkila, 2015). It has shown that fear after a disruptive event is more likely to lead to not only a lack of change and innovation, but also compromises at the community level.
Finally, this study sheds empirical light on the Bangladeshi and Pakistani apparel supply chains, demonstrating the difficult relationship between worker unrest and CSR practices. Following previous research on Bangladesh and Pakistan on the contrary ways workers obtain higher wages and benefits (Mishra, 2012;Munir et al., 2015), this study adds to Anner's (2018)

| Managerial implications for buyers and suppliers
While aligning with Anner's (2018) critique of buyers' pressure on supplier senior managers through low prices and high turn-around times, this study warns buyers of the contingencies of their aggressive and unethical pricing policies after the Rana Plaza incident. The interviews conducted in Bangladesh and in Pakistan are indicative of how buyers exert a paradoxical and dual pressure on suppliers to maintain apparel prices at minimum while simultaneously expecting them to widely engage with CSR in terms of more rights, entitlements, and better welfare to workers (Alexander, 2020;Foerstl et al., 2015). This paradoxical and dual pressure metastasized into market tensions that fueled market fear and anger in Bangladesh and Pakistan, and was well exemplified by Shakil-one of the supplier senior managers interviewed-when he shouted that buyers "are sucking our blood!" Although market fear and anger might, perhaps, unexpectedly alter the image of CSR practices as competitive tools to regain a good standing in certain contexts (e.g., Pakistan), lower prices and mounting costs would inevitably challenge the delivery of CSR practices for workers and create a divide between supplier senior managers. This divide can tarnish mutual relationships in the business community, as the concerns around worker unrest demonstrate. Arguably, lower prices might even encourage suppliers to adopt piecemeal and symbolic changes under the CSR rubric to gain visibility rather than provide transformative social upgrades for workers, such as fairer living wages and better welfare conditions.
While increased decisional freedom regarding CSR practices might cement trust, buyers should be aware that improving working conditions through better CSR practices requires complicity and a concerted effort through fairer and more ethical market conditions that factor in the costs of these practices. It might also decrease their perceived power asymmetry with suppliers and lead to a shift from relational contracting to value creation.

| Limitations and avenues for future research
This study has limitations that offer opportunities for future research.
Although fear and anger are among the most relevant discreet negative emotions (Coget et al., 2011;Lerner & Tiedens, 2006), humans are subjected to feelings that might differ by context (Lebel, 2017;Schieman, 2006). Supplier senior managers operating in other apparel supply chains might experience different negative or even positive emotions after the Rana Plaza incident in comparison with those in Bangladesh and in Pakistan. Future research would, thus, benefit from investigations on a broader theory of emotions applied to supplier senior managers after the incident in other developing countries' supply chains and how they correlate with CSR decisions.
The decision to focus on fear and anger as a unitary construct in this study was predicated upon the operationalization of fear and anger in the micro organization literature on negative emotions. Fear and anger are often believed to reinforce each other and be a consequence of each other (Gill & Burrow, 2018;Lebel, 2017;Onkila, 2015), especially after a disruptive event (Friedrich & Wüstenhagen, 2017). Despite their relatedness, it is salient to acknowledge that fear and anger can also be treated separately because their underlying mechanisms are not the same. It would therefore be helpful for researchers to provide a deeper examination of fear and anger as distinct emotions in future studies, and especially how they can exert influence on CSR decisions in supply chains without being necessarily in association with each other.
This study was centered on Rana Plaza as a critical incident that occurred in Bangladesh (Chowdhury, 2017;Donaghey & Reinecke, 2018;Fontana & Egels-Zandén, 2019). As a result, it led to social repercussions that were witnessed first hand by supplier senior managers in Bangladesh but not necessarily in Pakistan. The reason for exploring in a comparative study the effects of a disruptive event that occurred only in Bangladesh was substantiated by the interest to isolate negative emotions and observe their different influence on CSR decisions in two contexts having otherwise a very similar market and social structure. On the contrary, the authors concur that future research would benefit from focusing on the effects of one disruptive event on emotions in two or more contexts where it was experienced homogenously. This could perhaps highlight the existence of additional elements influencing CSR decision beyond emotions, including normative, cultural or regulatory factors.
Due to empirical difficulties, the study excluded the moral and normative dimensions of managers (e.g., their inner motivation and possibly intrinsic commitment toward alleviating workers' suffering).
Scholars are, thus, encouraged to evaluate the moral compass of supplier senior managers in apparel supply chains as the main mechanism behind their motivation and will. This evaluation will largely benefit the field because it can help gain insight about which elements to leverage to better frame CSR practices and improve labor conditions.
Similarly, it could provide more evidence against or in support of supplier senior managers being inherently unethical (Bird & Soundararajan, 2018;Jiang, 2009;Soundararajan et al., 2018).
Finally, this study adopted a multiple-country, comparative, and qualitative approach (Eisenhardt et al., 2016), which is essential to understanding fear and anger and explaining their correlation with organizational practices (Scherer, 2005). However, focusing on the apparel supply chain of one developing country and supplier senior managers from different upstream levels could convey a more heterogeneous understanding of fear and anger across that apparel supply chain. For instance, scholars could expand on this study by investigating the function of fear and anger among senior managers in decentralized second-tier suppliers as opposed to first-tier supplies.
Although the evidence on first-tier supplier senior managers is scarce, that of second-tier supplier senior managers is completely unexplored in the GVC literature. It could reveal new insights and help enable CSR practices to improve labor conditions at various upstream levels of the apparel supply chains.

ACKNOWLEDGEMENT
We the authors are particularly grateful to the supplier senior man- where he was a post-doctoral fellow and conducted most of this research. CSSI receives funding from Newmont Goldcorp Inc.

CONFLICT OF INTEREST
We the authors confirm that there are no financial and personal relationships with other people or organizations that could have inappropriately influenced our work. We the authors also confirm that the paper is our own and does not include any conflict of interest. Each supplier senior manager interviewed volunteered to disclose his knowledge.