Open strategizing on social media: A process model of emotional mechanisms and outcomes from un-orchestrated participation

This study extends the discussion on open strategizing by following the development of unlimited participation in an ex-ante city merger case where participation was not orchestrated. The findings unravel how the aggregation of emotional expressions on social media results in an escalating conflict, and initiates attempts to mitigate grounded in decision-makers ’ reflexiveness of social becoming(s). We theorize three emotional mechanisms – acceleration of emotional interaction, reinforcement of hostility as a discursive norm, and emotional empowerment – and outcomes originating from uncontrolled dynamics of inclusion and transparency. As such, we theorize open strategy as emerging organically on social media, and only later becoming deliberately orchestrated by strategists. Our findings have broader implications for understandings of emergent vs. orchestrated inclusion, individual vs. collective transparency, and open and closed decision-making in open strategizing. We also provide directions for managing inclusion to achieve positive outcomes in open strategy.


Introduction
Inclusion and transparency form key traits of open strategizing (Whittington et al., 2011;Seidl et al., 2019), as organizations increasingly recognize the various benefits of providing a wider array of actors with increased access to information, and including them in strategy formulation processes (Hautz et al., 2017).In city strategy-making in particular, the public play an active part in shaping strategy formulation through various types of formal activity, such as participation in fundraising, writing Op-eds, participation in petitions and demonstrations, joining NGOs and political parties, and voting (Quick and Feldman, 2011).As pointed out by open strategy scholars (Hautz et al., 2017;Luedicke et al., 2017;Malhotra et al., 2017), increased inclusion raises needs and expectations for access to and transparency of information.Thus, organizations around the world increasingly use online forums and social media to support stakeholder and citizen involvement in strategizing (Whittington et al., 2011).Although information technologies may be procured to managerial advantage (Whittington et al., 2011), the use of social media specifically "has the capacity to elicit new forms of meaning making for both positive and negative outcomes" (Knight and Tsoukas, 2019: 185, see also Glozer et al., 2019).
In this article, we focus on the emotional mechanisms involved in open strategizing, which become particularly salient on social media.Scholars argue that the effectiveness of open strategizing is dependent on whether the strategists can control the dysfunctional escalation of openness, and the resulting information and interest overload (Dobush et al., 2017;Hautz et al., 2017;Luedicke et al., potentially disruptive effects of social media, a trend in the "wider society in which strategists work" (Whittington et al., 2011).
As social media has entered human behavior, two affordances in particular, namely relational persistence and pervasive awareness, have been argued to fundamentally change the organizing of communities, such as collective action and public deliberation (Hampton, 2016).Social media affords persistent contact and interaction (Gibbs et al., 2013;Leonardi, 2018;Treem and Leonardi, 2012), as well as the articulation of associations to individuals, groups, or institutions over time, where "social ties have the potential to become enduring channels of communication" (Hampton, 2016: 110).Social media allows for direct involvement, that is, the sharing of particular worldviews with many others, as well as indirect and unintentional involvement through the visibility and exposure of content, and commitment of distant social ties (Hampton, 2016).
Previous studies on open strategizing have, for example, demonstrated how organizations can utilize online crowdsourcing software and social media to shape strategic content, by developing capabilities of reflexiveness that integrate external feedback into the organization (Baptista et al., 2017).In their examination of the literature on open strategy formulation, Malhotra et al. (2017) identify design principles that involve both process and technology features to utilize online crowds in overcoming knowledge gap risks.They note that although social media enables knowledge sharing related to a specific strategic issue, the current challenge in using online crowds is that the platforms are not designed to integrate knowledge-sharing activities.These examinations have, however, focused on how online platforms and social media features could be orchestrated for the purpose of strategizing.Our study focuses on how a strategic debate emerges on social media without being formally initiated by an organization.
When openness and transparency are key organizational traits grounded in democracy, public organizations not only deal with dilemmas of openness but also with the emergence of new practices.This involves the entrance of social media as the main site for deliberation that shifts traditional power dynamics between politicians and citizens.When actors engage in debates on social media, the medium allows for unlimited and uncontrolled interconnection and interaction irrespective of geographical borders (Hydle, 2015).Actors also continuously engage in consuming and co-producing content (Huang et al., 2013;Stieger et al., 2012), without waiting for the exchange of information and assignment of roles according to traditional norms of political decision-making processes (Mount et al., 2020).Furthermore, it has been suggested that when interacting online, actors are not necessarily aware of the embeddedness of the interaction in certain communities, and the intentional and unintentional consequences of communication online are not well understood (Leonardi and Vaast, 2017).
This has implications for how we understand open strategizing in public contexts, particularly when the strategic discussions of actors invested in decision-making move online, and there is persistent conflict among decision-makers with the public.First, there is limited knowledge on technology that cannot control participation.The previous research on open strategizing has investigated, for example, the use of specific organizational communication platforms to increase participation (Dobush and Kapeller, 2017;Malhotra et al., 2017;Mount et al., 2020;Plotnikova et al., 2020).The research has also examined mitigation practices, such as controlled agenda-setting and issue-framing (Malhotra et al., 2017;Luedicke et al., 2017), to deal with dysfunctional aspects of openness.Social media as a socio-material site for unlimited participation remains, however, understudied within the open strategy literature.Also, we do not know how dynamics of inclusion and transparency shape each other when openness is a key organizational trait; not orchestrated by, for example, top managers, and where no control exists over "the rules of the game of practices" (Vaara et al., 2019).
We suggest that open strategizing regarding city strategy-making on social media is an illustrative case of deliberate and emergent features of transparency and inclusion.

Social media and emotional expressions
While social media can facilitate the emergence of new material enablers that increase inclusion, transparency, and reflexiveness (Mount et al., 2020), there are risks, too.Actors' concerns can be voiced without the consideration of any emotional norms (Leonardi and Vaast, 2017;Lewis et al., 2014;Toubiana and Zietsma, 2017), as the interactive and open space allows them to express as well as respond to others' emotional expressions.In their recent essay, Knight and Tsoukas (2019) explored the role social media plays in providing a new way for organizational conflicts to emerge beyond organizational boundaries.The authors argued that the use of multi-modal modes of communication enables managers to sustain political dominance in politically and emotionally charged issues.These emotion-laden online influencing activities spurred increased participation (Bárberá-Tomas et al., 2019;Stieglitz and Dang-Xuan, 2013), and might also be directed at disciplining organizational behavior or stakeholder groups (Glozer et al., 2019;Toubiana and Zietsma, 2017).This would support the claim that emotions have a moral dimension and are grounded in the social control of what kind of behavior is considered "appropriate forms of emotional expressiveness given the setting and who else is present" (Samra-Fredericks, 2004: 1112).Online communication may, thus, result in an echo chamber where actors interact only with those who share their views, where their views are further reinforced, and alternatives silenced (Harel et al., 2020;Del Vicario et al., 2016;Leonardi et al., 2013).Vocal actors and actor groups can thus "undermine and restrict the participation of other groups," challenging subject positions in open strategizing (Splitter et al., 2021).
We suggest that in those types of organization, that is, public, which are "born open" (Hautz et al., 2017), and where inclusiveness is a prerequisite and extreme (Luedicke et al., 2017), there is a greater risk of suffering from the dysfunctional effects of openness.When public debates between citizens, and between citizens and politicians, move online, affordances of social media allow emotions to surface, and they trigger content sharing processes with other users (Serrano-Puche, 2021).Eberl et al. (2020) discovered that the more explicit the positive/negative language of politicians' Facebook posts, the more "Love/Angry click speech" reactions the posts received.Similar results were found in a study that examined to what extent anger and fear were related to distinct information-seeking and debate patterns (Wollebaek et al., 2019).The authors discovered that anger could reinforce echo chambers, if directed towards the confirmation of existing beliefs with like-minded people.
E.-L.Lundgren-Henriksson and V. Sorsa The risks of open strategizing online are evident in what has been termed "affective polarization" (Harel et al., 2020;Suarez-Estrada et al., 2022).It refers to discourse that instrumentalizes emotional expressions, such as anger, shame, and fear, against a political agenda to create a gap between "Us and Them".Citizens' animosity, and growing distrust towards the out-group, together with growing exposure to partisan news underly affective polarization (Serrano-Puche, 2021).Affective polarization reinforces collective action by strengthening the ties between likeminded people (Dolata and Schrape, 2016), and reduces the opportunity for civilized conversation with the opposition group (Harel et al., 2020).This has serious consequences for open strategizing.The mobilization of negative emotional expressions not only reinforces stereotypes around "the Other" and hostility towards those holding divergent views, but may also seriously challenge decision-makers' control over inclusion, the development of the agenda, and visibility of conversations in strategic debates (Haefliger et al., 2011;Stieger et al., 2012).
Yet, on some occasions in an offline space, when community anger is acknowledged rather than shield away from, managers have been able to engage in adversarial debates in ways that have restored inclusion (Feldman and Quick, 2009).Recent research has also looked at how angry cues offline, but also partly expressed on social media, spark the mobilization of collective anger and protest contributing to the emergence of strategic organizing (Kudesia, 2021).These studies suggested anger is not merely dysfunctional, but may facilitate collective sensemaking by generating new cues and reactions through which crowds and communities can organize themselves.It is not clear, though, how the acknowledgement of negative emotions and engagement in adversarial debates take place on social media.Emotional cues and reactions can spread even faster, and people can build on each other's emotional expressions differently than offline.Furthermore, the origin and outcomes of emotional expressions' aggregation in a crowd unrestrictedly debating strategic issues online remains unexplored.
We now turn to our empirical setting.

Research methodology
The ex-ante merger debate concerns two municipalities (Alpha and Beta) in a bilingual region, where Alpha is significantly smaller than Beta, and has a higher number of inhabitants with a minority identity.Our focus is on the division between actors who take part in the discursive coalitions in Alpha.These actors are politicians and other stakeholders, such as citizens, who adopt the competing standpoints of Pro and Anti merger, and tensions therefore emerge within the same municipality.The empirical material was collected in situ, enabling scrutiny of the twists and turns in the merger debate.The data comprise interactions on social media, interviews, publicly available texts, and observation of publicly available recordings (council meetings) (see Table 1).
The research process followed the principles of inductive qualitative research (Locke, 2001;Pratt et al., 2020), where concepts emerging from the data became the focus for further data collection and analysis.Thirty in-depth interviews with councilors served as a source to build an understanding of how politicians experienced the merger debate.At the start of the data collection process, while the interviews did not especially focus on the merger debate and how it played out online, the topic naturally emerged in many of the stories.Therefore, many of the respondents were interviewed twice to gain a more comprehensive view specifically on the role of social media in the merger debate.The interviews were not used primarily as a source to capture emotional expressions, rather to understand the actors' general viewpoints regarding the polarized public debate, as spectators or as participants.
The dynamics of emotional expressions were primarily captured through discursive analysis (Kouamé and Liu, 2021) of the publicly available and naturally occurring verbal and visual interactions on social media and in the public press.Key politicians' posts, from those who were for or against the merger and active on social media, were followed over time, as were the long and intense debates sparked by those posts between stakeholders in comment threads.We collected all the merger-related public posts from these politicians.In all, the social media material comprised more than 200 posts and over 8000 comments.The observation of video recordings of council meetings, where issues related to the merger were debated, acted as a source from which to elaborate on our emerging findings.Overall, we paid particular attention to how actors responded to and questioned each other, and hence how discursive contests emerged in situ across spaces, and shaped the overall media debate.
The analysis started by constructing an overall chronological case story on the merger debate, and delineating three chronological acts: the emergence of uncontrolled inclusion; the activation of defensive interaction from open strategists; and, the introduction of conciliatory interaction and controlled inclusion (see Fig. 1) (Langley et al., 2013).While the data collection was guided initially by our assumption that the political underpinnings of emotional expressions were significant, iteration between the literature and the ongoing collection of empirical material helped us specify our findings.Thus, the analysis proceeded in stages and followed the principles of live coding (Locke et al., 2016), where multiple and iterative rounds of coding underlie the abstraction of the findings.First, the lead author of the present study read through the interview transcripts and publicly available material, and articulated ideas for more in-depth exploration.Second, after discussion, we identified the two coalitions for an in-depth analysis that was refined through analogical reasoning utilizing the literature (Mantere and Ketokivi, 2013).The discourse of the two coalitions was examined to identify the key arguments in each of the acts and analyze the subtext of the negotiation and the goals of the two coalitions.The results of this stage of the analysis are presented in Table 2.
Next, our focus shifted from the content of the discourse to the verbal expression of emotions in the public material.We drew on Coupland et al. (2008) in that we identified emotional expressions as social performances where people articulated, to themselves and others, aspects of the merger.Following the discussion in the previous section, notions of the merger and what it enables and restricts were of central interest, as were notions of professionalism, appropriateness, and legitimacy.Consequently, we narrowed our empirical Fig. 1.The unfolding merger negotiations.
E.-L.Lundgren-Henriksson and V. Sorsa focus on how the actors experienced the contests as the merger debate intensified, and the reasons underlying participation in or withdrawal from the debate.This route emerged from the insights that actors' discursive behavior changed across physical and virtual spaces, and that the overall discursive climate, particularly on social media, was aggravated as the merger debate proceeded.Hence, we turned to the literature to explain the escalating conflict we observed.
Table 3 details the steps in the analysis process.Our coding of emotional expressions was inductive and highly iterative (see Table 4 in Appendix 1 for detailed illustrations).Based on our first, close reading of texts (Op-eds and social media), we developed our initial list of codes related to emotional expressions.We then turned to the literature on established coding schemes of emotions to guide us in our refinement of the identification of patterns of expressed emotions.
As we continued coding, first-order emotional expressions were grouped into theoretical emotion categories (prototypical

Table 2
Comparison of the coalitions' argumentation in the public display.

Overall subtext
The region continues to lack resources in national competition The democratic rights of the minority community are weakened nationally

Goal of the coalition
Growth -Merging constitutes an opportunity for the region and the minority community Securing independence and local democracy in Alpha -Merging constitutes a threat to the minority community Speed up the merger decision-making process Postpone the merger decision-making process

Key arguments Act 1
"Alpha council has elected the council board, which is in line with democratic practices" "Alpha council board composition tilted towards politicians in favor of the pro-merger coalition, which is not democratic" "We need growth, not stagnation" "Continued independence is not stagnation"

Key arguments Act 2
"The merger agreement has a future focus" "The agreement is a catastrophe for Alpha" "The pro-merger coalition is undermining democracy by rushing the process"

Key arguments Act 3
"Rise of a region" "Beta's financial situation is worrying -Citizens of Alpha will have to pay the debts of Beta" "Alpha's financial situation is worrying -Alpha needs the economic aid a merger brings" "The pro-merger coalition has influenced the merger process -The agreement is not democratically produced" "The anti-merger coalition has crossed the line -The debate does not benefit from personal attacks" "The debate does not benefit from personal attacks"

Emotional empowerment
Increasing spatial reach of discursive activity and defining group discursive interaction E.-L.Lundgren-Henriksson and V. Sorsa emotions), based on previous literature following the circumplex model (e.g., Russell, 1980;Scherer, 2005): fear, sadness, anger, joy (Shaver et al., 1987).This model allowed us to organize our first-order coding into emotional expressions based on two dimensions: positive or negative (i.e., hedonic valence), and high and low intensity (i.e., activation level) (Russell, 1980;Shaver et al., 1987).As Liu and Maitlis (2014), we found that the model allowed us to discover and explore nuances within the emotional prototypes, which can be denoted as (expressed) mood states (Grandey, 2008).For example, again following Liu and Maitlis (2014), coding guidelines on "anger" were adapted and used to capture both outrage and distrust.We distinguished between these expressions based on their intensity, that is, distrust being less intense than outrage (Scherer, 2005).Likewise, we followed coding guidelines on "joy" in terms of positive emotional expressions, yet treated expressed hope as less intense than enthusiasm.We used coding guidelines on "fear" in capturing worry, and "sadness" for dissatisfaction.
We followed recommendations on coding verbal and non-verbal cues related to the four major emotion categories in our data (Shaver et al., 1987) (see Appendix 1).We predominantly identified the emotional expressions through verbal cues in the texts (social media posts, comments, and Op-eds), through the direct use of emotion-related wordings and the tone and content in the context (Toubiana and Zietsma, 2017).Valence (i.e., positive or negative emotional orientation) was identified through the content explicitly or implicitly expressed.Intensity (i.e., low or high) was identified through the use of, for example, exclamation marks, capital letters, and swear words.
Despite our data being dominated by written and visual text, in recordings of council meetings, we also identified emotional expressions based on vocal expressions, such as emphasis on words, pauses, and changes in tone, complemented by non-verbal cues, such as facial expressions, and to some extent also body language.Since we realized that both coalitions adopted a multimodal approach on social media, particularly in later stages of the debate, we coded verbal and non-verbal cues in videos and visuals on Facebook.We identified positive emotional expressions by observing how actors were, for example, smiling, laughing, and gesticulating in videos.Visuals, such as memes, included negative emotional expressions, through, for example, pictures of war zones accompanied by text using satire to articulate the detrimental consequences of merging.
In all, we identified worry, dissatisfaction, outrage, and distrust as negative emotional expressions, and hope and enthusiasm as positive, capturing the heated debate and its turning points.Although we identified the negative emotional expressions across all chronological stages in the "messy" debate, and these should hence be treated as overlapping, particularly "stronger" negative emotional expressions, such as outrage and distrust, dominated the debate in the later stages.
We then focused on detecting what these emotional expressions accomplished, and accordingly how they could be considered as emotional mechanisms (Liu and Maitlis, 2014).We started looking for patterns of frequently occurring expressions of emotions, especially on social media, and shifts over time across the data.Also, how these expressions were discursively utilized by actors in the respective coalitions to resonate and mobilize support in the ongoing debate.Hence, we treat emotional mechanisms as the continuous aggregation of similar emotional expressions.We further investigated how emotional mechanisms seemed to be embedded in the expectations of, and demands for, emotional expressions within the local setting (Rafaeli and Sutton, 1987).That is to say, whether and how particular expressions violated norms, and what patterns of expression were reciprocated to "punish" or mitigate these expressions.Based on our emerging understanding that emotional expressions differed across public spaces and over the course of the debate, we turned to literature on the formation of space (Stephenson et al., 2020).Thereon, we sought to further explain the shifts in emotional mechanisms, for example, by considering their socio-material embeddedness (Balogun et al., 2014).Finally, we turned to the open strategy literature in theorizing how the identified emotional mechanisms could be approached in terms of inclusion and transparency.

Findings: From uncontrolled to controlled inclusion in open strategizing on social media
We present our findings chronologically, following the development of the merger debate, and illustrate the associated dynamics of inclusion and transparency in open strategizing on social media.

Act 1: The emergence of uncontrolled inclusion
In the print media the merger debate was initiated as Op-eds (increasing in 2017) authored by both citizens and politicians, directed at articulating the competing rationales for opinions on the merger, as well as establishing commitment thereto (see Table 2).The merger opinions were articulated to the public by expressing worry about the future, but also by informing the public about the competing merger opinions, actors' faulty statements, and merger rationales.For example, the pro-merger coalition's future image was assessed as utopian by the anti-merger coalition, which portrayed a bigger municipality resulting in "the death of centennial local history".The initiating merger discussions therefore emphasized extreme inappropriacy, that is, the realization of the merger entailed the risk of losing local services and the sacrifice of the minority community's democratic rights.Emotional expressions of dissatisfaction soon emerged: "The advocates of the merger are emphasizing a negative development in Alpha and only positive sides with a bigger municipality.Aim; disfavoring the present and promising gold and green forests."(Op-ed, anti-merger politician) "I don't know if the aim is to mislead the readers, since the truth is that the regions which we are competing with have merged municipalities to strengthen their positions."(Op-ed, pro-merger politician) The main platform on which the merger was contested was Facebook.In late 2017, politicians in the respective coalitions started to E.-L.Lundgren-Henriksson and V. Sorsa express their views on the merger in Facebook posts.This included increasing new entries of, and connections between, participating stakeholders (politicians, officials, and citizens), as the medium allows for unlimited and uncontrolled interconnection and interaction irrespective of geographical borders (Hydle, 2015;Mount et al., 2020).Actors continuously responded to and reciprocated emotional expressions and the debate soon turned ugly, as the rapid development of the merger debate and the increasing participation cannot be controlled (Stieger et al., 2012).In the comment sections, coalition actors discursively built emotional expressions, such as dissatisfaction towards each other, both by engaging in direct interactions with opponents and supporting likeminded participants in the debate: Anti-merger politician: So sad to see that those who promised to work for independence in the elections have now changed their opinion.
Citizen: Dishonest individuals have no place in politics, they prove it through their devious actions.
Both posts and comments entailed strong emotional expressions and revealed the conflict over the strategic decision-making process, which "includes disagreements about assignments of duties" (Jehn, 1997: 540) that manifest as personal back and forth attacks (Malhotra et al., 2017) on the duties of elected politicians to (not) realize the merger.Through these intersubjective interactions, stakeholders of the respective coalitions not only signaled but also strengthened their merger standpoint and association with likeminded actors.Phrases such as "fanatics of independence" or "fusion friends" when referring to the counter-coalition were particularly prevalent.These discursive activities collected input from others adopting the same viewpoint, simultaneously establishing commitment and participation in the debate.Constructing the opposing coalition through exaggerated emotional expressions, such as dissatisfaction, also consolidated in-group identification around their standpoint: Pro-merger politician: I can't understand how the friends of independence have completely lost it.Frustration bordering on hatred.
Other pro-merger politician responds: It's incomprehensible to read this [post of anti-merger politician] … if [name of antimerger politician] is spreading this in the villages, that's destructive.A language fight is the last thing we need.
Actors increasingly consumed and interpreted (the increasing amount of) merger-related information presented and discussed on social media, in line with their adopted merger standpoint.The actors are entwined with others in a particular practice world, a sociomaterial practice, which offers actors an orientation, understanding, and sense of meaning (Sandberg and Tsoukas, 2020).When acting within this world, that is, becoming immersed in "absorbed coping", actors are not paying deliberate attention to their actions; to a great extent they are driven by an anticipation of how the practice will play out.This points to the immanent features of interaction actors become absorbed in their routinized practices, infused with the situation, and spontaneously responding to it.In the increasingly heated debate, stakeholders continuously engaged in cycles of reciprocating discursive behavior simultaneously, enacting reality and constructing their merger standpoint as the only correct one (Orlikowski and Scott, 2008).This serves to further strengthen identification within the discursive coalitions ("us") and dis-identification between the groups ("them").This experienced distance between the coalitions was expressed in the interview accounts: "I've been involved in the debates and realized that you can't convince anyone.They've already made up their mind and joined one of the two camps.There's nothing you can say.And there's nothing that can convince me either."(Respondent 3) "You cannot distinguish between opinion and person, it's like 'okay you have this opinion so then you are dishonest, you're a bad person'."(Respondent 10).
"We need transparency and debates in an open society … but the debate is so emotional and goes on at parallel levels, then it's difficult … The disagreement consumes all energy."(Respondent 5).
In sum, this stage reveals how the discursive coalitions emerged and strengthened.Social media offered a place for individuals to interact and share their experiences, discuss and debate (Leonardi and Vaast, 2017), and engage in joint sensemaking (Ellison et al., 2011), to such an extent that views on issues became shared (Leonardi, 2018).Actors identified and discursively engaged with stakeholders expressing the same merger opinion regardless of geographical locations, not necessarily aware of the community in which they were immersed when joining conversations (Leonardi and Vaast, 2017).At the same time, reinforced expressive statements on the other discursive coalition excluded their worldview and merger opinion, creating boundaries to what kind of views (and who) belonged to the "outside".

Act 2: The activation of defensive interaction from open strategists
In 2018, when the drafts of the merger agreement, including the structure of the future city organization, became public, the contentious conflict escalated (Malhotra et al., 2017).An uneven distribution of power (cf.Luedicke et al., 2017) grew in the debate as politicians were increasingly subject to accusations, and finding it hard to dominate the discussion.The number of Op-eds in the print media was also increasing, and many started to resemble the social media debate inasmuch as personal attacks were presented to the public.Emotional expressions displaying distrust increased.Debating the rationales for the merger opinions blurred with how citizens and politicians appeared to each other in the debate: E.-L.Lundgren-Henriksson and V. Sorsa "Based on statistics, it can only get worse for Alpha […] it's incomprehensible that a group of politicians led by [name of promerger politician] are allowed to destroy our well-managed independent municipality."(Op-ed, Citizens and anti-merger politicians) An exchange of Op-eds emerged in the press, also directly linking to discursive activities on Facebook, to express outrage related to others' participation in the debate: "I'm shocked by the attack [name of anti-merger citizen] in the newspaper [date].I've kept a record of all attacks that have been made on Facebook!I think the independence agitators would gain from developing positive arguments for their cause instead of personal attacks."(Op-ed, pro-merger politician) The number of participants in the debate on Facebook, especially citizens, increased rapidly when the merger agreement draft became the center of debate.Actors from other regions joined, even from other countries, expressing the need to voice their opinions on the merger.In particular, the anti-merger coalition's publicly displayed interaction intensified, when both key politicians and citizens expressed a fear that the minority community would have no decision-making power in the future city.These intense exchanges conveyed outrage regarding the merger process, and distrust towards leading politicians, who were accused of imposing the merger.This was also articulated in an emotional performance at a council meeting in early 2019, a speech that was verbally displayed in a Facebook post: "The merger agreement must be a joke.A bad joke!The worst agreement ever.It's a capitulation."(Anti-merger politician) This illustrates how social media enables defensive reactions to spontaneously emerge (cf.Kudesia, 2021) also from politicians, that is, strategists.At that stage in our case, politicians, and other stakeholders such as city officials, who had until then officially adopted a neutral view on the merger, started to play an active part in the debate online.Key events mobilizing more heated discussions emerged when these actors, who were positioned to make credible claims, publicly criticized the merger agreement draft on Facebook, displaying "negotiations" in public that would otherwise have unfolded behind closed doors.This resulted in a more evident and growing polarization.At that stage in the debate, the discursive climate was rapidly aggravating, and reached breaking point at the end of 2018.
In late 2018, one of a politician's posts received over 500 comments.Negative emotional expressions continued to strengthen, such as outrage: Anti-merger politician to pro-merger politician: I completely agree with [name of anti-merger politician] that this bullying must stop -Why are you '***ing' with [name of anti-merger politician]?Is it because [name] doesn't want to go to Beta? Pro-merger politician replies: According to you, if the 'wrong person' is criticized then it's'***ing with him/her.'In that case you are '***ing with' a lot of people yourself.
Due to social media's distinct characteristic of easily maintaining interaction and contact (Gibbs et al., 2013), as well as its open nature fostering unintentional voyeurism and exposure to content of also distant social ties (Hampton, 2016), actors become both involved and observers on social media (Leonardi, 2018).Especially the public visibility of expressed emotions on the part of officials and politicians, regarded as boundaryless social media behavior, mobilized discursive commitment from stakeholders who had previously only been observers of the debate.This was visible in comments on social media such as "now it's time to react," "It was not my intention to comment here but I felt a need to," "It doesn't feel right to write on this thread, but …" or "I avoid debating on social media, but I'm participating in this debate to defend our democratic system."Expressed emotions were increasingly deployed to discipline discursive activity.Stakeholders joined, or continuously rejoined, the debate to "correct errors" in "manipulated facts", such as through expressed distrust: Pro-merger politician to anti-merger politician: I'm ashamed for your sake.You're accusing the report of being biased.Then you do the same thing.
Pro-merger politician: This is unjustifiable.Negotiations should not be made via media and not on Facebook.Unless you have the intention to completely ruin it.
The rapid accumulation of increasing social media posts and comment threads made the escalating conflict between politicians, and between politicians and citizens, more visible to the public.At that point, participants in both discursive coalitions were increasingly starting to question the discursive behavior of all stakeholders taking part in the debate.Key stakeholders' discursive behavior on social media, such as sharing links, liking posts, or commenting posts was unexpected.Further, it was articulated in the interviews and public debate as extremely inappropriate and unethical, given the public nature of the online interaction: "As a public decision-maker, maybe we should think twice before entering a heated discussion."(Respondent 12)."I'm disappointed at some experienced politicians taking part in the debate, and partly heating it up.I've encouraged them several times to distance themselves from it.If you're talking about democracy and justice, then you must stand up when it's violated.No one does that now […] throughout the debate some politicians have taken advantage of the fact that you can accuse others of being undemocratic, which has frightened people away from participating in the debate.This is dangerous from the perspective of democracy."(Respondent 4) Although still embedded in the hostile discursive climate of the debate, participants were simultaneously negotiating the boundaries for acceptable discursive behavior, thus engaging in discursive "boundary work".It encompassed articulating, responding E.-L.Lundgren-Henriksson and V. Sorsa to, and "defending" accusations of being "undemocratic".This displacement is activated by temporary breakdowns (Sandberg and Tsoukas, 2011) when actors perceive deviations and boundary crossing, giving rise to reflection of the actual socio-material practice that becomes the object of deliberation (Sandberg and Tsoukas, 2020).This "temporal and spatial" demarcation aimed to start directing the emotional expressions back into a civilized discursive realm, and separating the prevailing negativity and personal attacks from the debate on the merger.We argue that this emerging "deliberativeness" was demonstrated by politicians exiting the debate, or restricting their posting and commenting on social media, as well as to some extent also publishing Op-eds.This stemmed from reflections and worry that the aggressive public debates were influencing public opinion, the future public image of the region and the minority community, and even the political profession.As such, the (political) decision-makers' assessment of social media as a space for public debates initiated reflection on their future commitment not only to the organization (Luedicke et al., 2017) but also to the entire profession: "If it continues like this, it might be difficult to recruit people into local politics in the future.Who wants to be part of these types of dirty debate?"(Respondent 4) In sum, specifically in Act 2, disciplining discursive activity simultaneously intensified the already heated social media discussion, as it was grounded in contentious demarcations of what type of discursive activities belong in public spaces and open strategizing.This was preceded by temporary breakdowns and active reflections accumulating into a major breakdown (Jarzabkowski et al., 2019), a collective "revelation" that the debate was derailed and in need of active control.

Act 3: The introduction of conciliatory interaction and controlled inclusion
In its later stages, the debate rapidly became more extensive on social media, and to some extent also in the printed media and council meetings.More people continuously "self-selected" (e.g., Stieger et al., 2012) themselves into the debate to exert influence in favor of their preferred outcome prior to the referendum and final decision-making.In January 2019, the merger agreement was made public.The time leading up to the referendum was characterized by intense discursive work on influencing public opinion, and after it on influencing the politicians that had not yet taken a clear stance on the merger.During those months, the merger debate was visible on a daily basis in the local media and on social media.
The increasing deliberation of politicians in 2019 (print media, council meetings) stemmed from a developed reflexivity regarding the need to clearly demarcate legitimate discourse, and to restore faith in public decision-makers and the democratic discussion.The debate was consequently directed towards "forming separations or spatial distinctions" (Stephenson et al., 2020: 812), to (re)direct the discursive activities and the increasingly emotional nature.Emotional expressions were modified to display hope for the community's future, particularly in the pro-merger group, and strongly articulated in daily Op-eds: "Conflicts emerging from discussions are natural.Change evokes strong emotions and fear of losing something valuable.We need to remember that the foundation of communication is trust.We hope that we can develop a consensus, where trust and belief in the future are building blocks for the merger.We all need to put the personal attacks aside and accept a changing society."(Op-ed, pro-merger politician) In the last council meeting before the vote and merger decision was taken, emotional expressions of hope unfolded.These elaborated social performances connected past and future, emphasizing possibilities for community identity were (not) the merger realized.These also incorporated references to the derailed debate across the other discursive spaces: "The merger evokes emotions, and the problem is that things have been said that one regrets.It is important that we, who will continue working together, should say 'I'm sorry' […] The municipality is the future of our citizensthe role of the municipality is not the same as of a city."(Pro-independence politician) "Today we set the future agenda.We can develop and flourish only by becoming a city."(Pro-merger politician) The debate revolved around facts and figures related to the potential merger, and subsequently what is "right and wrong".The heated crowd was increasingly demanding to be "listened to" and emotionally claim further inclusion, such as to be considered in the final decision-making (Luedicke et al., 2017).At the same time, the decision-makers engaged in "process structuring" (Mount et al., 2020), by imposing "rules" that directed the content of communication to influence public opinion but also soothe the heated debate.Organized and targeted discursive activities emerged alongside the heated debate on Facebook, which aimed to communicate, encourage, and shape a different type of inclusivity in the debate, and to reach out to a larger mass of people before the referendum.This included clear, organized links between Op-eds and Facebook, and the introduction of multimodality (videos and visuals).These modified patterns of emotional expression and discursive activity were particularly visible in the pro-merger group, where politicians implemented a campaign on social media and in the local press.As such, the pro-merger coalition's discursive activity intentionally grew in intensity.
" 'The villages will be destroyed!'[referring to the anti-merger discourse] I mean, where will they go?We have deliberately decided not to use such language in the discourse in our group."(Respondent 8) "We are trying to focus on the positive aspects of the merger, not personal attacks but facts and objectivity […] the most important thing is that as many as possible vote in the referendum.It will facilitate making the final decision."(Respondent 3) Control over the strategic discussion was also exercised by directing how the agenda was manifested in personal accounts of both E.-L.Lundgren-Henriksson and V. Sorsa decision-makers and citizens.Interviews with key authorities, citizens, and politicians (verbal text and videos) were posted on social media daily, focusing on articulating and discussing the potential benefits of the merger, displaying hope.In videos, both pro-merger politicians and citizens provided their perspectives on the future by drawing on personal experiences, all organized around the same topics of discussion, which included the possibilities that growth would bring for the community: "The word 'merger' evokes negative emotions, but I take a positive view of it as I have long personal experience of what a merger means in the business world [..] we need to see the light and look forward."(Video interview with pro-merger citizen, campaign of pro-merger politicians) "We can win the battle between regions if we merge and become a bigger city.It's not about what I gain, it's about what coming generations will gain, what my children and grandchildren will gain."(Video interview with pro-merger politician, campaign of pro-merger politicians) The emotional expression of hope and enthusiasm also included vivid illustrations (text and images) of the future community, such as children and grandchildren, relying on the present actions of both citizens and politicians.The strategic use of those emotional expressions also discursively marked a distance from the negatively-oriented emotional expressions.Invitations to participate in the debate were constructed differently to those in previous stages in the derailed debate: Campaign of pro-merger politicians: Tomorrow might be a historical day!The day we plant the seeds of something that will grow and prosper.Whatever the result [referendum] we will work for a strong region!These types of discursive encouragement aimed to foster a constructive sharing of views and ideas from participating stakeholders, to share perspectives with the decision-makers assessing their final merger standpoint, and communicated attending to the concerns of inclusivity expressed by the crowd, in turn empowering the strategists: Campaign of pro-merger politicians: Our region is successful due to our ability to look forward and create structures that secure our future.How do you see the region 10 years from now?What does everyday life look like?Please share your thoughts!However, expressed hope was ultimately not shared to the extent that would have fully mitigated the derailed debate.The antimerger coalition was also organizing the discourse and emotional expressions multimodally to expand inclusivity and spatial reach; memes expressing satire through manipulated pictures of local pro-merger politicians were posted daily in a public Facebook group.In particular, expressed distrust continued to be echoed within the coalition as the merger debate moved closer to the referendum: Anti-merger politician: It's incomprehensible that the merger has been initiated and powered by frenzy in Alpha.Very nice that [name of anti-merger politician] spoke out.There's more to sayjust wait.
Other anti-merger politician: It's very important that [name of anti-merger politician] informs us, so that the citizens have a chance to form the correct opinion.
Other anti-merger politician: Good that you open our eyes, we who have not been part of the negotiations.
In sum, the public merger debate quickly escalated into personal discrediting through expressed negative emotions, to the point  where legitimate discursive activity was breached.The online interactions were clearly different from face-to-face interactions (Gibbs et al., 2013), and the emotion-laden messages on Facebook tended to engage more people (Bárberá-Tomás et al., 2019;Stieglitz and Dang-Xuan, 2013;Toubiana and Zietsma, 2017).Social media also enabled the shaping of expressions (Huang et al., 2013).The modification of discursive activity and an increase in organized discourse accordingly took shape in Act 3, and was further aided by the introduction of targeted videos on social media.

Discussion
We now turn to discussing our findings, drawing on the prior open strategy (e.g., Hautz et al., 2017) and social media (e.g., Treem and Leonardi, 2012) research.We direct particular attention to the emergence of three emotional mechanisms: acceleration of emotional interaction, reinforcement of hostility as a discursive norm, and emotional empowerment.We examine how these shape stakeholders' discursive participation in discussing and negotiating strategic issues, and identify key outcomes from social media emerging as the primary yet unintentional forum for open strategizing (see Fig. 2).
First, as actors build on each other's posts and comments on social media, including emotional expressions, the continuously growing content supports the development of common ground (Treem and Leonardi, 2012;Leonardi and Vaast, 2017), and formation of collective identities (Luedicke et al., 2017), around the opposing strategic perspectives.This reinforces collective action (Dolata and Schrape, 2016).Group membership and the respective perceptions of reality and "truth" are accordingly strengthened, enabled by growing close social ties (Hampton, 2016) and signaling their association and relationships with likeminded actors (Treem and Leonardi, 2012) within the debate.Likewise, the rapid increase in displays of negative emotions from both within and outside of actors' network domains, stimulates more engagement and discursive activity on social media (Eberl et al., 2020;Wollebaek et al., 2019).That in turn reinforces emotional expressions and fosters the emergence of an emotional echo chamber (Toubiana and Zietsma, 2017).
The rapid and uncontrolled inclusion in the debate is shaped by stakeholders defending or accusing individuals rather than the strategic idea (cf.Luedicke et al., 2017).Simultaneously, they engage in reproducing different knowledge within the two distinct groups (cf.Malhotra et al., 2017) regarding the strategic issue (i.e., merger reality) and the stakeholders taking part in debating (i.e., "us" and "them").Hence, discursively connecting with others in the heated debate shapes boundaries between merger opinions and individuals, and the contest over subject positionings (Splitter et al., 2021); that is, decision-makers and citizens.Taken together, we theorize this first emotional mechanism characterized by a low degree of reflexiveness as acceleration of emotional interaction.
The case findings suggest that when actors are building on each other's content (information presented on social media and offline) with a low degree of reflexiveness in a debate characterized by affective polarization (Serrano-Puche, 2021;Suarez-Estrada et al., 2022), rapid and uncontrolled inclusion contribute with distorting strategic perspectives, and it becomes impossible for stakeholders, including the decision-makers involved, to discern the actual strategic issues from the debate.For example, the widespread occurrence of negative attacks impedes a collective focus on exchanging views, on matters such as strategic challenges that could constructively inform politicians' final decision-making.Actors within the discursive coalitions struggle to take each other's view into account, only promoting the in-group's view (cf.Malhotra et al., 2017).This adds to what Hautz et al. (2017: 302) argue vis-à-vis the dilemma of disclosure: "The further away the respective audiences are from the context in which the information originated, the more likely that the information will be interpreted in the light of a different interpretive context resulting in conflicting interpretations." Moreover, from the perspective of decision-makers (i.e., politicians), the growing content alters their motivation to contribute further to the debate (Treem and Leonardi, 2012)."Boundary work" manifests, as actors increasingly start to contentiously demarcate the boundaries of discursive activities on social media, and to appropriate emotional expressions according to the context and audience (Samra-Fredericks, 2004).The normalization of hostility on social media emerges as "affective polarization can reduce the opportunity for civil discourse" (Harel et al., 2020: 7).As others' realities are made visible (Leonardi and Vaast, 2017;Treem and Leonardi, 2012), and actors continuously engage in consuming and co-producing content on social media (Huang et al., 2013), hostility becomes institutionalized as a norm in the debate (Dolata and Schrape, 2016).We theorize this second emotional mechanism as reinforcement of hostility as a discursive norm.
In addition, the increasing level of reflexiveness emerging from a heated debate showcases how social media not only affords but also constrains the discursive participation of decision-makers (Leonardi and Vaast, 2017).A high degree of social transparency online enables actors in the debate to observe who the participants are and how they are communicating (Leonardi and Vaast, 2017), fostering the collective assessment of decision-makers debating in a public discursive space, and ultimately resulting in a perceived challenge for legitimacy and inclusion.When actors' discursive participation in the polarized debate started to adopt traces of "deliberativeness", it originated from both "watching" the discussion climate and the sense of "being watched" (Hampton, 2016).Despite actors taking part in the heated debate by disciplining others, they demonstrated an increasing reflexiveness (Baptista et al., 2017) in their discursive behavior as a collective group; they started to reflect on the future social positioning of the political profession (Mantere and Vaara, 2008).The heated public debate appeared to undermine and threaten the public's view of the societal and democratic relevance of the political profession, as well as its legitimacy (Hautz et al., 2017;Whittington et al., 2011).As such, the findings demonstrate how deep reflection (Mantere and Vaara 2008) on professional role identity demotivates (cf.Luedicke et al., 2017) and impairs decision-makers' further participation (Stieger et al., 2012) in strategic discussions online and even offline.
As in previous conceptualizations of the process dilemmas of open strategy (Hautz et al., 2017), the findings demonstrate how social media enables the crowd debating a strategic issue to grow stronger, and to direct and dominate strategy negotiations (Stieger et al., 2012) between decision-makers, as well as between decision-makers and citizens.The introduction of rules of the game has been argued to ameliorate this power asymmetry (Malhotra et al., 2017).The case findings, however, demonstrate that when re-occurring E.-L.Lundgren-Henriksson and V. Sorsa discussion topics revolve around questioning the decision-making legitimacy of decision-makers, and they in turn engage in defending their legitimacy, the introduction of discursive rules (i.e., through disciplining others) contributes to aggravating the process dilemma of open strategy.
Social media allows for a high degree of editability and strategic tailoring of messages, including directing what type of and how personal information is made visible (Leonardi and Vaast, 2017;Treem and Leonardi, 2012).The case findings show how decision-makers actively utilize social media features to redirect strategic discussions, overcome knowledge distortion, and redirect knowledge flows, by inviting the crowd to share knowledge with each other and with decision-makers differently (cf.Malhotra et al., 2017).The third emotional mechanism, emotional empowerment, is thus characterized by a high degree of reflexiveness, and facilitates an active (re-)involvement of stakeholders in the strategic discussions.Social media enables asymmetries between the distribution and control of power to emerge (Leonardi and Vaast, 2017), but also their mitigation through the exercise of social influence (Dolata and Schrape, 2016).
Thus, the new course of the strategic discussions and controlled inclusion restores power relations between citizens and politicians, and accordingly mitigates power asymmetries, enabled by emotional empowerment.This demonstrates how the deliberate mobilization and orchestration of emotional expressions also becomes a way for decision-makers to reclaim authoritative control over strategic discussions (Malhotra et al., 2017), as well as over decision-making legitimacy.This offers implications for the dilemma of escalation.Hautz et al. (2017) argue that the open nature of strategic information flows might result in challenges to restrict openness, as the crowd demands access to further information and inclusion in decision-making, possibly resulting in negative attacks if the crowd's expectations are not met.The findings show how selecting the strategic issues to be discussed, presenting concrete and factual arguments, and emotionally calling to a past and future shared cultural identity, aim to foster constructive idea sharing and attend to a crowd's hostile requests for inclusion.

Contributions and conclusions
We have explored a case that encompasses unrestricted and uncontrolled participation in a pre-merger debate in public discursive space(s).We argue that it is particularly illustrative of open strategizing, and how open strategy can be dysfunctional under certain circumstances (Hautz et al., 2017).Our findings enabled us to theorize emotional mechanisms and outcomes originating in dynamics of inclusion and transparency, especially as shaped by social media (Haefliger et al., 2011).The emerging open strategy literature is relatively silent on the role and effect of technology that stimulates uncontrolled discursive participation (Vaara et al., 2019), as well as the emergence of affective polarization (Serrano-Puche, 2021;Suarez-Estrada et al., 2022).We theorize open strategy as emerging from a crowd (Kudesia, 2021) and organically on social media, and that only later becomes deliberately orchestrated by strategists.
Our study informs the literature on 1) open strategy and social media, and 2) open strategy and emotions, demonstrating how the entrance of social media as the main site for debate offers an arena to dispute and (re-)negotiate power relations between decisionmakers and the crowd (Mount et al., 2020;Splitter et al., 2021).First, recognizing the dynamics between spontaneous emotional expressions and strategic intentionality (Kudesia, 2021) advances insights on emergent vs. orchestrated inclusion.The prior research suggests that the effectiveness of open strategizing is dependent on whether the strategists can control the dysfunctional escalation of openness, and the resulting information and interest overload (Dobush et al., 2017;Hautz et al., 2017;Luedicke et al., 2017;Malhotra et al., 2017).Our study shows that social media shapes open strategizing by engaging participants in uncontrolled discussions that shift the focus from strategic issues to emotionally provocative content.Social media enables disembodied voicing of views and expression of emotions that produce a sentimentalizing effect on public discourse (Serrano-Puche, 2021).Here, the challenge for open strategizing is not as much in the information and interest overload as in the emotional expressions reinforcing echo chambers and affective polarization (Harel et al., 2020), sustaining a contentious conflict (Malhotra et al., 2017) between decision-makers and the crowd.Conversely, our study also demonstrates how inclusion can be orchestrated by decision-makers through the rhetorical use of emotions (Samra-Fredericks, 2004), and how it serves as a medium to mitigate this particular conflict by regaining control over agenda setting and the framing of strategic issues (Luedicke et al., 2017;Malhotra et al., 2017).Thus, the identification of emotional mechanisms advances insights on practices of inclusion, which have previously been theorized as "activities that involve making connections among people, across issues and over time to engage in an ongoing stream of issues" (Mack and Szulanski, 2017), to include also deliberate and emergent features of emotional expressions.
Second, we theorize the role of emotional mechanisms in individual vs. collective transparency.The persisting and visible nature of discursive activities online seemed not only to stimulate participation as actors continuously grew social ties within the debate (Hampton, 2016), but also reflective activities such as the adjustment of actors' (collective) discursive participation.This suggests that discursive participation in online strategizing can be shaped by an increasing (collective) level of reflexiveness.We argue that this advances insights on discursive boundary work (Stephenson et al., 2020) to include the generation of reflection during wider participation.Also, insights on affective polarization as a source for a collective "awakening" regarding "dysfunctional escalation", which can spark controlled activities aiming to create and sustain a present and future community of interacting stakeholders (Hautz et al., 2017).
Moreover, in line with Baptista et al. (2017), the case demonstrates that social media provides a discursive and participatory space for commenting on and contributing to open strategizing.Although they argue for the direct causality between commenting on social media and shaping strategic content through an organization's developed "reflexiveness", we cannot establish the causality of the derailed debate on the merger outcome.In the case, reflexiveness is not a required organizational capability to foster openness and inclusion, but a capability to mitigate relational and societal breaches, and to (re-)establish the legitimacy of decision-makers as strategists (i.e., politicians) and advocating their societal responsibilities (Hautz et al., 2017).The dynamics of discursive participation in open strategizing were grounded in actors' reflections on their subjectivities and positionings within present and future strategic deliberations (Baptista et al., 2017).The findings thus show that when open strategizing becomes characterized by affective polarization visible to the public, it has an influence on practitioners' commitment as they struggle to maintain their subject positions (Splitter et al., 2021), possibly even stimulating reflections on continuing their service to a profession or organization (Jehn, 1997).Building on Splitter et al. (2021: 27), we argue that participation in open strategizing that emerges on social media entails "an ongoing struggle", at not only the inter-but also intra-subjective level of "those who try to maintain their subject position as strategists".Hence, the findings offer a nuanced view on the continuous interplay between individual and collective aspects of transparency (Hautz et al., 2017;Whittington et al., 2011).
Third, the identified emotional mechanisms and outcomes have implications for the dynamics between closed vs. open decisionmaking.The prior research on open strategizing suggests that participation can be restricted (Dobusch and Müller-Seitz, 2014), and authoritative control re-established after collecting input from the crowd (Malhotra et al., 2017).Our findings challenge these linear conceptualizations by demonstrating how divergence of viewpoints and persistent conflict among decision-makers and the public might be triggered by social media sustained over time (cf.Luedicke et al., 2017).Open strategizing is thus an ongoing process (Jarzabkowski et al., 2019) where actors participate in strategic discussions online without being formally "invited".This has implications for the time and resources invested in open strategy negotiations, as the crowd's input cannot easily be discerned or taken into consideration by decision-makers, since relevant information is missing (Luedicke et al., 2017).That is due to affective polarization and continuously ongoing negotiations that compromise speed and flexibility in the strategy formulation process (Hautz et al., 2017).Nevertheless, our findings emphasize, in line with Feldman and Quick (2009), the importance of approaching inclusion as "never complete" and as "the continuous development of a community of participation" (p.139).They refer to "resourcing" -"actions that create potential resources and that link potential resources with frameworks in which they can be used" (Feldman and Quick, 2009, p. 140), and suggest that citizens' emotions, such as community anger, can constitute a strategic resource, if recognized and deliberately engaged with.Our study suggests that community anger online cannot be fully controlled due to social media affordances.Therefore, resourcing demands online an increased level of reflexiveness, and as our findings show, deliberate and collective discursive structuring to support inclusion and decision-makers' legitimacy.
Last, the findings theorize discursive activities aimed at resolving the tensions of openness and transparency (Baptista et al., 2017) as increasingly dynamic, by the identification of multimodal emotional expressions online as well as offline.As the findings show, taking control of both inclusion and transparency can manifest through decision-makers (re-)introducing a strategic agenda that aim to emotionally "speak to" a shared community identity.This broadens understanding on how orchestrated inclusion of identity-related issues becomes a way to mitigate the "dysfunctional escalation of openness" (Hautz et al., 2017).This type of mitigating practice can take the form of "manipulated" videos and images on social media, embodiment in offline performances (i.e., council meetings), and be directed as much to the crowd as to decision-makers themselves.
We invite future research to investigate our theorized emotional mechanisms and outcomes of open strategizing, in order to unravel how these play out across different organizations, that is, other contingencies, as well as space(s) for debate.Which practices offer inclusivity and transparency on social media in decision-making processes, without challenging consensus building?Over time, which discursive practices "close" as opposed to "open" inclusivity in public debates, and when are these applied?Despite social media facilitating citizens' expression of their voice, and becoming active and participating in the democratic dialogue and strategizing, our findings show that the consequences for open strategizing and decision-makers can be unfavorable when debates are uncontrolled and unfold publicly.Future research could explore in depth the possibilities for uncontrolled inclusion in open strategizing on social media to generate strategic input for integration into strategy formulation.
In conclusion, our findings offer guidance for public managers.First, the findings recognize the need for explicit discursive action plans in public organizations, including discursive "codes of conduct" for participation in informal social media debates.Also, the emergent nature of affective online debates guides managers to actively seek cues for negative emotional escalation, to inhibit affective polarization.This requires active monitoring of uncontrolled discussions online, as well as planting "cues" in the form of emotional expressions that can guide collective interpretation (cf.Kudesia, 2021).Second, although our findings largely focus on the "dark sides" of social media inclusion, uncontrolled inclusion can be recognized as a strategic opportunity.Acknowledgement of and engagement with active citizens collectively addressing on social media public issues and policy-making, offers public managers opportunities to find new perspectives.It also guides them to regard emotional expressions and divergent perspectives as a strategic resource (Feldman and Quick, 2009) with the potential for creative strategy input (Kudesia, 2021).

Declaration of competing interest
None.An earlier version of this manuscript was presented at the Annual Academy of Management Meeting in Boston, 2019.The authors wish to kindly thank Associate Editor Julia Hautz and two anonymous reviewers for their valuable feedback during the review process, and Linda Rouleau for her helpful feedback in early stages of developing the manuscript.

Fig. 2 .
Fig. 2. A process model of emotional mechanisms and outcomes of open strategizing on social media.

Table 1
Research material.

Table 3
Data structure.Building on the debate by commenting on in-and out-group social media posts and Op-eds Expressing strong reactions of anger or shock • Expressing declaration that someone is responsible for a fault and/ or cannot be relied upon • Increasing and targeted personal attacks in Op-eds and social media related to "questionable debating" and spreading of "false realities" • Displaying crossing of discursive boundaries by others e.g., by linking to other discursive spaces • Expressing that someone or something does not meet expectations • Expressing fear of actual or potential problems • Expressing merger rationales and commitment in Op-eds and on social media • • Expressing expectation for a particular thing to happen • Expressing a positive outlook • Discursive separation from previous stages of the debate by deliberatively connecting and rhetorically organizing discursive spaces • Introduction of organized groups and multimodality on social media Modifying discursive activity through hope and enthusiasm