From Support to Mutiny: Shifting Legitimacy Judgments

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/276182964 From Support to Mutiny: Shifting Legitimacy Judgments and Emotional Reactions Impacting the Implementation of Radical Change Article in The Academy of Management Journal · Decemb...

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/276182964 From Support to Mutiny: Shifting Legitimacy Judgments and Emotional Reactions Impacting the Implementation of Radical Change Article in The Academy of Management Journal · December 2014 DOI: 10.5465/amj.2012.0074 CITATIONS READS 192 4,200 3 authors, including: Quy Nguyen Huy Kevin G. Corley INSEAD Imperial College London 77 PUBLICATIONS 5,463 CITATIONS 68 PUBLICATIONS 19,060 CITATIONS SEE PROFILE SEE PROFILE Some of the authors of this publication are also working on these related projects: Strategy implementation in emerging economies View project The (Changing) Nature of Strategic Change View project All content following this page was uploaded by Quy Nguyen Huy on 09 June 2015. The user has requested enhancement of the downloaded file. 娀 Academy of Management Journal 2014, Vol. 57, No. 6, 1650–1680. http://dx.doi.org/10.5465/amj.2012.0074 FROM SUPPORT TO MUTINY: SHIFTING LEGITIMACY JUDGMENTS AND EMOTIONAL REACTIONS IMPACTING THE IMPLEMENTATION OF RADICAL CHANGE QUY NGUYEN HUY INSEAD KEVIN G. CORLEY Arizona State University MATTHEW S. KRAATZ University of Illinois Based on a three-year inductive study of one organization’s implementation of radical organizational change, we examine the critical role played by middle managers’ judgments of the legitimacy of their top managers as change agents. Our analysis revealed middle managers’ shifting judgments of the change agents’ legitimacy that arose with their emotional reactions and produced rising resistance to the change effort. Our inductive model illustrates the dynamic, relational, and iterative relation- ships among change recipients’ legitimacy judgments of change agents and arising emotional reactions in various phases of planned change, which explain recipients’ emergent resistance to the change effort. Our model allows us to contribute to theory on radical organizational change, resistance to change, and legitimacy judgments. Organizational scholars have long recognized the principle, emerge from anywhere inside the organ- severe challenges involved in organizational trans- ization (e.g., Plowman, Baker, Beck, Kulkarni, So- formation efforts and the necessary role that indi- lansky, & Travis, 2007); however, top managers vidual agents play within them (Amis, Slack, & (TMs) have been the conventional focus—and for un- Hinings, 2004; Hinings & Greenwood, 1996; Huy, derstandable reasons. These actors play a readily vis- 2002; Singh, House, & Tucker, 1986). Planned rad- ible and public role in many change initiatives and ical organizational change (PROC), which funda- are often hired for the express purpose of transform- mentally alters the power structure, culture, rou- ing failing organizations. In addition, they appear to tines, and strategy of the entire organization, often possess many of the resources that are most necessary appears to be the only option available in dire for this task, such as formal decision authority, con- circumstances (Miller & Friesen, 1980; Tushman & trol over resources, and centrality (e.g., Denis, Lang- Romanelli, 1985). As such, scholars have generally ley, & Cazale, 1996; Romanelli & Tushman, 1994). treated radical change as an inherently volitional Over the last two decades, scholars have exam- phenomenon and placed change agents at the cen- ined many different instances of radical change and ter of their theories and empirical research (Bar- presented ample evidence that supports an agentic tunek, 1984; Pettigrew, 1985). These agents can, in view of the process. They have also provided many useful insights about change agents’ role within This research was funded by the Social Science and this process and some prescribed requirements for Human Research Council of Canada and INSEAD Re- success in the job (e.g., Tushman & Romanelli, search. We would like to thank Alex Bitektine, Christiane 1985; Wiersema & Bantel, 1992). Despite its many Demers, Ann Langley, Henry Mintzberg, Leigh Tost, important contributions, a closer examination of Frances Westley, Spencer Harrison, as well as audiences this literature reveals at least one fairly systematic at Harvard Business School, Wharton Business School, IE Business School, and IESE for helpful comments on and seemingly significant oversight: existing theory previous versions of this article. We would also like to and research seem to have mainly focused on the acknowledge the exemplary editorial guidance we re- early stages of the radical change process and given ceived from Jason Colquitt. insufficient attention to the challenges that change 1650 Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express written permission. Users may print, download, or email articles for individual use only. 2014 Huy, Corley, and Kraatz 1651 agents face during its later stages— especially im- tern of events clearly affirmed our orienting beliefs plementation. This striking imbalance is likely due about the critical importance of implementation, and, in part to the opacity and complexity of implemen- as we will later suggest, represents a contribution in tation processes, especially in light of the difficulty its own right. Its main function, however, was to researchers have in gaining deep access into organ- provide a foundation for our inductive analyses to izations to study the PROC process closely and explain this critical stage of the process and develop longitudinally (Van de Ven, 1992; Ford, Ford, & a grounded theory about the aspects that change D’Amelio, 2008). It may also reflect, however, a agents (continuously) influence within it. Our data more basic misunderstanding about the nature of revealed that MMs frequently seemed to analyze their these laborious processes, or, even, a bias toward superiors’ statements and actions as if they were look- an overly strategic perspective of them. ing for clues about their motivations, intentions, and Regardless of the causes, though, there does seem capabilities—that is, engaging in continuous judg- to be a clear need (and opportunity) for research ment of Tekco’s senior change agents. These judg- and theory that delve into the implementation of ments were interlaced with emotional reactions that radical change and illuminate the role that change often influenced their action–response to those judg- agents play therein. This need becomes all the more ments. Over time, these perceptions seemed to co- apparent when one considers implementation as a alesce into more holistic, person-centered judg- critical dimension within the larger process of rad- ments— ones that had important implications for the ical change and the inherently agentic nature of larger change process. Specifically, we found that this process itself. Radical change efforts frequently MMs’ beliefs about their bosses were closely associ- fail as a result of implementation problems (Beer & ated with their beliefs about the larger change effort Nohria, 2000; Nag, Corley, & Gioia, 2007), yet there and the nature of their participation therein. The ex- is no logical reason to believe that change agents istence of this relationship became most apparent in should suddenly lose their ability to facilitate the the final days of Tekco’s radical change initiative, process as it passes into this critical stage. This when a final affront from the TMT caused MMs’ paper reports a study designed to examine this emotions to boil over and to stage an open revolt that puzzle and meet the need (and opportunity) for collectively rejected the authority of their bosses. more theory in this area. This led to the TMT’s rapid departure and brought an Our qualitative, inductive study followed a rad- abrupt end to the change effort. ical change initiative undertaken at “Tekco,” a These findings were not easily interpretable large and historically successful company that fell through the lens of existing theory on PROC. In into a deep performance crisis in the wake of de- addition to under-examining implementation, rad- regulation and associated changes in its competi- ical change scholars have not investigated deeply tive and technological environments. Given our enough the more general problem of change agents general interest in implementation and executive obtaining and maintaining support from change change agents, we followed Tekco’s change effort participants. We came to realize that our findings over its full three-year course and conducted mul- could be explained with a clearer focus when we tiple waves of interviews with its new top manage- turned to the large, multidisciplinary literature on ment team (TMT) (which had been recently hired legitimacy. Scholars dating back to Max Weber from outside), as well as with a large number of the have recognized that power stands in constant need company’s middle managers (MMs). We ap- of legitimation, and also emphasized that the for- proached these MMs as if they were the “lieuten- mer is apt to quickly disappear in the wake of the ants” who would execute the TMT’s orders, and the latter (as with political uprisings and mutinies). “lynchpins” of the larger radical change process Legitimacy scholars have also assembled a large (two metaphors common in the literature— e.g., Ba- body of theory and evidence that illuminate the logun & Johnson, 2004; Huy, 2002). various sources of legitimacy and hypothesize Over the course of our time at Tekco, we ob- about the process through which it is amassed— served a fairly remarkable turn of events. Within a and depleted (e.g., Suchman, 1995; Tost, 2011; Ty- short time after their arrival, the company’s new ler, 2006). The inductively generated theory of TMT successfully formulated a plan for change and change agent legitimacy that we present in this rallied substantial internal support. This plan soon paper emerged as we delved into this literature and lost momentum, however, and ultimately failed brought its various insights to bear on our qualita- during its implementation phase. This overall pat- tive data about Tekco’s PROC efforts. 1652 Academy of Management Journal December Our study’s main theoretical contribution is to tunek, 1984; Pettigrew, 1985). While agents of highlight the importance of change agent legiti- PROC can, in principle, emerge from many places macy within the process of PROC, which ulti- in and around organizations (Battilana, 2006; Plow- mately provides a relatively detailed mechanism to man et al., 2007), TMs often appear to be the most explain shifting resistance to change via evolving suitable candidates for the job. These individuals legitimacy judgments. A key theoretical insight occupy positions that provide many of the material concerns the multiple and shifting bases of change and symbolic resources most necessary to the task agent legitimacy. Drawing from the legitimacy lit- of radical transformation, including the formal au- erature and our data, we identify several aspects of thority to alter structure, reallocate capital, hire and legitimacy judgment content and explain why they fire personnel, and publicly proselytize for cultural are apt to shift over the course of a PROC effort. change (Finkelstein & Hambrick, 1996; Tushman & Evolving judgments of change agent legitimacy pro- Romanelli, 1985; Wiersema & Bantel, 1992). It is vide a much-needed theoretical mechanism to sup- not surprising, then, that a significant portion of the port a dynamic (rather than static) and pluralistic large and theoretically diverse literature on PROC (rather than single-sided) explanation of shifting re- has focused on the role that top management sistance to change (Ford et al., 2008). Another impor- change agents play in initiating and directing it tant insight concerns the holistic and actor-centered (Battilana, Leca, & Boxenbaum, 2009; Kraatz & nature of these judgments. Though judgments of Moore, 2002; Miller, 1991; Pettigrew, 1987; Ro- change agents may involve various types of content, manelli & Tushman, 1994). we posit that these typically congeal into a summary Nearly all forms of major change pose implemen- judgment of legitimacy that has significant influence tation challenges and require some degree of sup- on change participants’ behaviors. port and participation at lower levels (Sonenshein, Additionally, our emergent insights add empirical 2010). These challenges are particularly severe in texture to emerging conceptual work on legitimacy the case of PROC, which involves simultaneous judgments (Bitektine, 2011; Tost, 2011). Specifically, change in multiple core organizational elements. we empirically demonstrate that legitimacy judg- Moreover, external constituencies such as inves- ments of human agents exist and can change rap- tors, regulatory agencies, and customers hold high idly, in part because they can be highly emotion- expectations for rapid performance improvement laden (in contrast to legitimacy judgments of more (e.g., Amis et al., 2004; Romanelli & Tushman, impersonal forms and structures that tend to be 1994). All these factors impose severe constraints slow-changing and unemotional). Emotional reac- on change agents facing both high time pressure tions can trigger fast and abrupt changes in legiti- and reduced resources to address challenging ten- macy judgments and influence the information that sions (e.g., Huy, 2002; Sastry, 1997); failure to ad- becomes noticeable to evaluators. These insights dress these tensions could exacerbate organization- extend the nascent literature on legitimacy judg- al decline and even cause organizational mortality ment (which is dominantly cognitive) with an af- (Singh et al., 1986). fective dimension, showing how both cognition Intriguingly enough, the PROC literature has and emotion can co-arise and influence evaluators’ tended to assume that TMs’ authority to lead subsequent interpretation and behaviors. Our study change is largely unproblematic and that they typ- thus bridges several key literatures—radical organ- ically perform a beneficial change agent role. Tush- izational change, resistance to change, institution- man and Romanelli (1985: 173–180), for instance, alism, and emotional reactions to change—with a contend that “only executive leadership can medi- pluralistic and relational perspective to provide ate between forces for convergence and forces for deep insight into the implementation of PROC. change” and “implement the set of discontinuous changes” inherent in radical change efforts. Even though the literature on MMs has documented their THEORETICAL GROUNDING active contribution to innovation in incremental or continuous change contexts (Kanter, 1983; Floyd & Planned Radical Organizational Change Wooldridge, 1992), the PROC literature tends to The PROC process is generally characterized by de-emphasize the role of MMs and to portray them sudden and intense change actions that fundamen- in a self-effacing role (for an exception, see Huy, tally disturb various groups’ roles, identities, and 2002). Likewise, most normative models of strategy interests that have co-existed for a long time (Bar- tend to accord MMs a supporting role at best 2014 Huy, Corley, and Kraatz 1653 (Shrivastava, 1986); executives are advised to re- cution challenges faced by TMs in enacting these duce equivocality and establish control systems seeming straightforward prescriptions, and, more (Simons, 1994) so that MMs can comply with and importantly, their interactions with and depen- act on executives’ clear directives. dence on MMs for assistance in implementing rad- Coincidentally, the resistance to change litera- ical change. ture has tended to assume a favorable view of TM change agents and treats resistance to change by Resistance to Change lower-level employees as dysfunctional (Ford et al., 2008). In fact, resistance to change is often seen as In light of the complexities of PROC and the problematic to PROC (Hinings & Greenwood, fertile conditions for the emergence of resistance to 1996). Thus, this stream of research tends to focus PROC (e.g., Bartunek, 1984; Ford et al., 2008; Furst on what change agents can do to overcome resis- & Cable, 2008), it should not be surprising that tance. Relatedly, resistors tend to be construed as research on resistance to change has blossomed individuals with constrained psychological or so- over the years. This literature has explored various cial attributes, or motivated by narrow self-interests strategies that managers can use to reduce em- that inhibit them from seeing and accepting the ployee resistance to organizational change, such as reasons for and benefits of the proposed change. sanction, persuasion, participation, or communica- Implicit in this literature is a largely static, one- tion (e.g., Kotter & Schlesinger, 1979; Nutt, 1986). sided (agent-centric) and judgmental view of resis- Unfortunately, after decades of research, findings tance to change. about their effectiveness remain inconclusive (see Likewise, the prescriptive PROC literature has Bartunek et al., 2011; Furst & Cable, 2008). For suggested relatively simple, generic prescriptions instance, inviting employees to participate in plan- to aid change agents, such as creating a sense of ning a change has been found to increase employee urgency and a compelling vision, performing pro- support for change (e.g., Coyle-Shapiro, 1999) and cedural and interactional justice, participation and to reduce support for change (e.g., Bruhn, Zajac, & cooptation of lower levels, and project leader cred- Al-Kazemi, 2001). Similarly, using sanctions and ibility and trustworthiness to overcome resistance edicts to force employee support for a change has to change (e.g., Bartunek, Balogun, & Do, 2011; been effective in some cases (Poole, Gioia, & Gray, Kotter, 1995). As intuitively commonsensical as 1989) and ineffective in others (e.g., Nutt, 1986). these prescriptions may appear, empirical research As many scholars reviewing this literature have has underexplored and even questioned how feasi- noted (Bartunek et al., 2011; Ford et al., 2008; Pid- ble or realistic it is for change agents to act success- erit, 2000), research on resistance to change has fully on these prescriptions, particularly in the con- tended to take (a) a favorable view of change and text of PROC (Furst & Cable, 2008). change agents and (b) a dominantly individualistic Because of challenges inherent to PROC, success and cognitive view of change resistors. Indeed, this often depends in part on employees voluntarily literature has tended to take the unquestioned cooperating with TMs to realize radical change. stance that change is good and resistance to change Adherence to the spirit of the change goals, rather is bad (Huy & Mintzberg, 2003), and, thus, that than just the letter, is necessary to overcome un- change agents occupy a legitimate and beneficial foreseen complications (Huy, 1999). Since discre- organizational role. Consequently, this stream of tionary cooperation, rather than mechanical com- research takes the implicit stance that resistance to pliance, is necessary to creatively address change is dysfunctional and explores what change difficulties that emerge during the implementation agents can do to overcome it. Second, and relat- process (Amason, 1996), even a modest amount of edly, resistors tend to be construed as individuals resistance or passive apathy can harm the success with constrained psychological or social attributes. of PROC. Studies have shown that MMs often func- Thus, they resist change because of personal cog- tion as “linking pins” who play a critical role in nitive and social limitations such as cognitive ri- interpreting and implementing top-down changes gidity, conflicting schemas, low openness to and fostering discretionary cooperation (e.g., Floyd change, risk aversion, and protection of self-inter- & Wooldridge, 1992; Huy, 2002). Thus, MMs can ests (see Ford et al., 2008, and Piderit, 2000, for serve as a decisive force in “making or breaking” overviews). radical change efforts. Unfortunately, few empiri- Ford et al. (2008) have challenged this traditional cal studies have taken an in-depth look at the exe- line of thinking and argued for a relational perspec- 1654 Academy of Management Journal December tive on resistance to change (see Bartunek, Rous- such judgments is the decision to accept existing seau, Rudolph, & DePalma, 2006), one that advan- power structures and obey managerial directives. tages neither change agent nor change recipient, When people believe that their superiors are enti- but, instead, focuses on their interactions. In this tled to their positions, they will also generally feel view, the presence of resistance to change is neither obligated to follow their orders and requests (Sel- inherently good nor bad, nor is it solely to be znick, 1969; Tyler, 1997; Weber, 1978). blamed on either change recipients or agents. The Recent conceptual advances in the literature on key is to understand how the reciprocal actions of legitimacy judgments (Bitektine, 2011; Tost, 2011) both agents and recipients work together to foster have stressed the importance of looking at the con- or inhibit resistance to change. Employees may re- tent of legitimacy judgments; that is, “the substan- spond differently to similar managerial actions de- tive perceptions and beliefs that underlie the judg- pending on how they interpret the change agents’ ment of an entity as legitimate or illegitimate” motives. Thus, some scholars have started examin- (Tost, 2011: 687). Synthesizing the large and frag- ing how the quality of interpersonal manager– em- mented literature on legitimacy content, Tost ployee relationships could influence whether em- (2011: 693– 694) argues that there are three main ployees judge the information conveyed by their dimensions of content underlying active or “eval- manager as supportive and credible, or as manipu- uative” legitimacy judgments: (1) instrumental, (2) lative and selfish (e.g., Ferris & Judge, 1991; Furst & relational, and (3) moral. Instrumental legitimacy is Cable, 2008). How employees make positive or neg- present when the entity (in our study, the TM ative attributions about leaders’ motives, or experi- change agent) is “perceived to facilitate the indi- ence positive or negative emotional reactions to- vidual’s or group’s attempts to reach self-defined or ward their bosses, could impact their behaviors internalized goals or outcomes” such as “percep- toward the proposed changes, irrespective of tions related to the effectiveness, efficiency, or util- whether these proposals are intrinsically beneficial ity of the entity.” Relational legitimacy exists when or harmful. The resistance to change literature has the entity is “perceived to affirm the social identity tended to confound resistance to the change con- and self-worth of individuals or social groups and tent (what is to be changed) with resistance to to ensure that individuals or social groups are change agency (leaders or agents of change) (see treated with dignity and respect and receive out- Ford et al., 2008). Thus, there is a need to distin- comes commensurate with their entitlements,” guish resistance to the change agent from resistance such as perceptions of “fairness, benevolence, or to the proposed change content. Whether employ- communality.” Finally, “an entity is perceived as ees accept a controversial change might depend on legitimate on moral grounds when it is perceived to the extent to which they evaluate the change con- be consistent with the evaluator’s moral and ethical tent and/or the change agent as legitimate or not. values.” She also notes that these three dimensions Legitimacy ample, “credibility” has been defined as the quality of being believable or trustworthy. This concept has re- Legitimacy is widely recognized as a critical so- ceived very little scholarly attention relative to legiti- cial phenomenon, and scholars across the social macy. Because legitimacy also portrays dimensions of sciences have evinced a longstanding concern with believability or trustworthiness, we thus focus on the understanding its nature, origins, and conse- more encompassing legitimacy concept. Moreover, a few quences (Beetham, 1991; Jost & Major, 2001; John- studies have looked at how top teams experience de- son, Dowd & Ridgeway, 2006; Deephouse & Such- clined credibility due to their inability to achieve polit- man, 2008). Integrating the vast institutional and ical support or substantive change outcome, but in the psychological literatures on the topic (e.g., Such- eyes of constituencies who have power equal or superior man, 1995; Tyler, 1997), Tost (2011: 688) defines to that of the top team, such as the board of directors, rather than subordinating MMs (see Denis, Lamothe, & legitimacy as “the judgment that an entity is appro- Langley, 2001; Denis et al., 1996). As for “fairness,” Tost priate for its context.”1 One critical outcome of (2011: 690) argues against some scholars’ tendency to conflate legitimacy with fairness, which is only one di- mension of the content that underlies legitimacy judg- 1 In the fragmented literature on organizational ments. Other dimensions of legitimacy exist, and fairness change, we note concepts that are close to “legitimacy,” can be an antecedent or outcome of legitimacy judgment such as credibility, trustworthiness, or fairness. For ex- (see Tost, 2011). 2014 Huy, Corley, and Kraatz 1655 are not mutually exclusive, in that they may over- through which to explore resistance to change and lap and that an entity could be evaluated simulta- agent–recipient interactions. In particular, it opens neously on all three dimensions or some subset of a window into factors that are external to the or- the dimensions. ganization, and provides a way to integrate macro, In addition to making active legitimacy judg- institutional considerations alongside micro, social ments based on these three dimensions, Tost (2011: psychological insights (e.g., Tost, 2011). It also avoids 695– 696) also suggests that people often assess le- the managerial bias that is evident in much of the gitimacy in a more “passive mode.” These judg- prior research on resistance to change (Ford et al., ments occur through some combination of two 2008). When the relationship between change agents other proposed cognitive processes. In the first, and change recipients is approached from a legiti- people use “validity cues” as cognitive shortcuts macy perspective, managers are not inevitably the and base their own judgments entirely on the en- “good guys” and resisting employees are not neces- dorsements of other observers. In the second, they sarily irrational, selfish, or obstructionist (Bitektine, simply accept entities that display appropriate 2011; Tost, 2011). Finally, the legitimacy framing also symbols and conform to cultural expectations. Be- invites needed attention to the role of interpersonal cause people tend to approach judgment tasks in dynamics and emotional reactions (both of which ways that minimize effortful cognitive processing, have received insufficient attention in the extant this more passive mode is expected to be operative PROC literature—Huy, 1999; Lazarus, 1991). much of the time (Lieberman, 2003). This form of legitimacy judgment is also present in situations Emotional Reactions where an entity is accepted merely because it be- longs to a generic category that is culturally “taken Emotion refers to a feeling state with an identi- for granted” as good, familiar, or non-problematic fied cause or target that can be expressed verbally (irrespective of its behaviors, outputs, or demon- or non-verbally (Elfenbein, 2007). The question of strated competencies—Bitektine, 2011). what is an emotion and what is a “borderline” Legitimacy has long been recognized as a vital emotion is still debated by emotion scholars resource for power-holders looking to develop and through various terms such as anger, excitement, maintain authority and extract high-quality com- hope, compassion, frustration, disappointment, or pliance from their subordinates (Beetham, 1991; surprise (for an elaborate discussion, please see Tyler, 2006; Zelditch, 2001). In its absence, Lazarus, 1991: 82– 83); thus, we use the term “emo- would-be authorities are unable to elicit voluntary tional reactions” to include both emotions and bor- cooperation from their subjects, and are, instead, derline emotions. Lazarus’s (1991, 1993) emotion forced to rely on costly and self-limiting coercive theory suggests that people typically experience tactics. In more extreme circumstances, the loss of emotional reactions as they evaluate the signifi- legitimacy can destabilize existing structures and cance of an event in relation to their own goals and lead to a wholesale loss of power itself (as in mu- concerns. If they appraise the consequence as ben- tinies and political uprisings). Though previous eficial, pleasant feelings arise. They experience un- theory and research on PROC has said very little pleasant feelings if they appraise the consequence about the legitimacy of radical change agents, its as (potentially) harmful. Thus, to the extent that potential role within the radical change process people appraise the legitimacy of their superiors is not difficult to recognize. Attention to change who have an important say about the nature of their agent legitimacy seems all the more important jobs or who lead change that is potentially impor- given the significant problems that change agents tant to the realization of people’s important goals are likely to face in gaining and maintaining it. and values, engaging in legitimacy judgment can Because these individuals must generally ask their arouse strong emotional reactions. subordinates to make significant and often disrup- Emotional reactions often generate a change in tive changes in their identities and routines, and to readiness to act that prepares people to take action accept significant sacrifices on behalf of the organ- (Frijda, 1996). People determine a potential action ization, their legitimacy is likely to be rather prob- response as they evaluate their own abilities to deal lematic and subject to ongoing scrutiny. with the event. If they determine they have ade- Another key benefit of bringing legitimacy theory quate resources to deal with the event, they are to bear on PROC is that it provides a more encom- more likely to respond actively. Otherwise, they passing, nuanced, and ideologically balanced lens may adopt a passive/avoidance approach, which 1656 Academy of Management Journal December could be interpreted as a form of resistance to mental change. The government, acting in response change. As emotional reactions can impact both to changes in technology and the political environ- thinking and behavior (Elfenbein, 2007), they could ment, also made a decision to deregulate the indus- influence subsequent legitimacy judgments and re- try. This deregulation paralleled changes that took sistance to change. Research has under-explored place in many other industries and countries dur- how exactly emotional reactions emerge during the ing the 1980s and 1990s as market logics gained implementation of PROC, or how they interact with increasing influence. As a result of these converg- legitimacy judgments and resistance to change to ing technological and ideological changes, Tekco help steer the course of PROC implementation. Our faced more new competitors, including both de inductive study seeks to explore these critical gaps novo firms and foreign technology companies. in our understanding. These contextual changes also exerted pressure on Tekco’s established culture. The company had historically been engineering dominated and in- METHODS ward focused. A number of government officials Research Setting and major customers felt that Tekco had not been responsive enough in terms of innovation and pric- Our data collection efforts focused on radical ing, while some of its executives were becoming change at “Tekco,” a large information technology too arrogant, yet risk averse. These new pressures (IT) company that enjoyed a dominant market po- manifested themselves in the company’s declining sition for more than 50 years.2 Starting as a small market share and profits, and also subjected it to venture more than a century ago, by the 1990s, harsher criticism and scrutiny from investors and Tekco had become a large provider of IT and com- securities analysts. munication services, with more than 50,000 full- In response, the board of directors appointed a time employees and a market value of more than new CEO from outside of the company—John Max- $12 billion dollars. Over time, its growing market well, who had held several senior management po- power attracted the attention of governmental reg- sitions with high-technology firms—and directed ulatory agencies that required the quasi-monopoly to him to turn Tekco around. Maxwell began by mak- provide universal and affordable service in the mar- ing sweeping changes to the TMT, splitting the kets it dominated. The company’s strength relied on a monolithic bureaucratic structure into multiple highly skilled engineering staff that provided com- business units with profit-and-loss accountability, plex, reliable, and state-of-the-art services. The com- changing the company’s incentive system, reduc- pany enjoyed decades of revenue and profit growth, ing its cost structure by shedding 25% of the work- and could thus provide employees with generous force, and recruiting large numbers of new manag- company benefits, life job security, and well-estab- ers to quickly develop new sales and marketing lished career development paths. skills for competitive markets (see Table 1 for a While the company had been historically profit- timeline of Tekco’s change efforts). able and stable, it faced an increasingly threatening To enact these changes, the new TMT enlisted an competitive, technological, and institutional envi- initial group of about 40 veteran senior MMs in the ronment in the years leading up to our study. By formulation of radical change plans to generate the late 1990s, advancements in computing power what specifically should be changed to restore prof- and IT posed a severe threat to Tekco’s established its. MMs are generally important in organizational business model and suggested a need for funda- transformation processes (e.g., Balogun & Johnson, 2004; Huy, 2002), and occupy a particularly impor- tant role in the case of Tekco. At the request of 2 The names of the organization and its members, eth- these MMs, TMs contracted a prestigious external nic origins, geographical locations, and specific technol- consulting firm, “Binary,” famous for quantitative ogies have been disguised to protect the confidentiality analysis. Binary consultants helped MMs perform of this publicly traded organization. The exact names of change programs have also been modified, although we extensive quantitative analyses and bring external tried to convey the spirit of the disguised names. In the benchmarking information to them, so as to deter- same spirit, some numbers and calendar dates have been mine the best areas for work reengineering. This altered. These adjustments in the presentation of the data initial group of championing MMs then enlisted do not affect how the data have been interpreted to ex- more than 500 other MMs across a wide array of plain the proposed theoretical concepts. work units—selected because of their work experi- 2014 Huy, Corley, and Kraatz 1657 TABLE 1 cluded the first author having unfettered access to Key Events in Tekco’s CT Program observe meetings and interact with employees, we Year 1 Market share: 76% conducted both formal and informal conversations January John Maxwell becomes CEO and starts replacing with employees at all levels of the company. For- senior executive team. mal interviews consisted of semi-structured discus- March Strategy formulation phase begins. sions about the general change effort and instances Select group of MMs and consulting firm to analyze and quantify proposed change projects. of change management specific to the informant; December Complete new senior executive team in place. interviews lasted for about an hour. Following the Year 2 Market share: 70% tenets of purposeful sampling (Lincoln & Guba, January [Data collection begins.] 1985), these interviews often ended with suggestions March Strategy implementation phase begins. for at least two other employees considered influen- Announcement of CT plan to reduce workforce by 13,000 in parallel with 150 change projects tial in one or several aspects of the change effort. Of championed by MMs. particular interest were those who might have a dif- June Announcement of terms of voluntary separation ferent viewpoint, so as to help maximize diversity program and support mechanisms. Downsizing and depth of perspectives, although our techniques begins. also sought confirmatory data (Charmaz, 2006). Year 3 Market share: 64% June Major decline in customer service known Perspectives were included in our analysis when throughout the company. provided by at least two informants (Laumann & December Implementation challenges have caused Pappi, 1976). concerns about major delays and This process resulted in the first author inter- underperformance of various major change viewing the incumbent CEO 3 times, the two pre- projects, in turn causing concerns that financial and customer service improvement targets vious CEOs once each, 12 executives 26 times, and might not be met. 114 MMs3 192 times, in addition to many lower- Year 4 Market share: 59% level employees and union officials. About half of August Change evaluation phase begins. MMs interviewed belonged to the change champi- Implementation challenges continue to surface, oning group, and the other half belonged to the causing concerns that originally forecasted change recipient group (although this categoriza- financial targets will not be met. TMT then decrees involuntary layoff of 3,000 people. tion is a rough one in practice, because a MM could September CEO Maxwell resigns. TMT dissolved. New CEO be a change agent of a particular change project appointed. while being the recipient of another change proj- ect). To get a more general perspective, we typically ence and high influence with their respective peer interviewed senior MMs who supervised other ju- groups. The goal was to propose various change nior MMs. As Eisenhardt and Graebner (2007) sug- initiatives that would allow the company to in- gest, interviewing informants from diverse groups crease revenues and cut costs quickly, with a target and perspectives reduces the risk that all these economic payback of three years as a key criterion informants would engage in biased, convergent ret- for obtaining funding. MMs suggested about 250 rospective sensemaking and/or impression man- change projects to improve Tekco’s operations and agement. The real-time data collection used in this profit, of which about 150 were funded. These proj- study also mitigated rationalization biases. Re- ects were collectively referred to as the Corporate search in real time is recommended for longitudi- Transformation (CT) program and involved reengi- nal research because it allows researchers to see neering work processes and the heavy use of IT (for events in a way that is closer to that actually expe- more details on these CT projects, see Huy, 2002, rienced by the participants, and that is sensitive to 2011). To fund these projects, the company set current activity and concerns. aside $1.2 billion of restructuring costs over the In addition to these formal interviews, the first next three years, with a goal to achieve a pre-tax author conducted informal conversations with annual cash flow of $700 million by December 31, 2006. 3 MMs are two levels below the CEO and one level above first-line supervisors. There are many levels of Data Collection middle management at Tekco, so one senior MM in the line groups could be in charge of 2,000 –5,000 frontline As part of a three-year research project examin- workers. TMs include the CEO and the level of execu- ing Tekco’s three-year radical change, which in- tives above MMs. 1658 Academy of Management Journal December close to 200 employees. These informal discussions agent or the change effort. Building from these first- were useful in corroborating what he saw and order codes, we then coded for similarities and heard in the formal interviews and observations, as differences within and across our informant groups well as in triangulating various processes of to detect conceptual patterns (Glaser & Strauss, change. The informal conversations also allowed 1967), as well as with the extant literature to ensure him to screen potential new informants for addi- that themes previously discussed in the literature tional formal interviews who had been suggested were correctly attributed (e.g., legitimacy—Tost, by previous informants; those informal conversa- 2011; resistance to change—Ford et al., 2008; emo- tions that raised new insights, or potentially diver- tional reactions to change—Huy, 2002) and new gent viewpoints led to requests for formal, recorded themes were properly specified. interviews. Finally, the first author also gathered One of those new themes was the presence of internal documents related to change project status, various emotional reactions to the change and ac- employee surveys, and internal HR efforts. These tions of the top team. Following prior empirical secondary data sources, as well as the results of examinations of emotional reactions to change observations and the perusal of confidential archi- (e.g., Huy, 2002), we initially relied on the “cir- val data, allowed the triangulation of findings from cumplex” model of emotions (Larsen & Diener, diverse sources to build stronger interpretations 1992) to explore the wide range of emotional reac- (Charmaz, 2006). tions people can experience during change. Ac- cording to this model, emotions share two basic dimensions: one dimension reflects the hedonic Data Analysis valence (pleasant– unpleasant) while the second re- Initially, narratives of the change effort were de- fers to the intensity of arousal (high versus low veloped from the perspective of both TMs and intensity). Together, the four dimensions capture MMs. Key dates and milestones were placed into almost the full range of emotional reactions (Bartel flow diagrams so that the dynamics of TM and MM & Saavedra, 2000). Thus, the hybrid category of actions and interactions could be better modeled pleasant, high-intensity emotional reactions in- (Langley, 1999; Pentland, 1999). Initially, these di- cludes enthusiasm and excitement, while pleasant, agrams looked very complex, because we tried not low-intensity emotional reactions include calm to discard too prematurely any factor that might and comfort. Unpleasant, high-intensity emotional turn out to be consequential and that might be reactions include anger, anxiety, and fear, while repeated across narratives (pattern matching). Two unpleasant, low-intensity emotional reactions in- consistent themes that began to emerge from the clude disappointment, shame, or dejection. MM change narratives were (1) the waning accep- Beyond noting explicit emotion terms such as tance and support for the TMT over time and (2) a “afraid” or “excited” for coding emotional reac- corresponding increase in descriptions of resis- tions, we also identified implicit emotional reac- tance to change. As we delved into the literature to tions by relying on Lazarus’ (1991) core relational understand these emerging themes, it became ap- themes and cognitive appraisal theory of emotion parent that, while scholars understand each phe- (e.g., Smith & Ellsworth, 1985). These scholars ar- nomenon individually, we have almost no theory gue that the experiences of specific emotions (e.g., on their interrelationship. anger, disappointment) are reliably associated with We then began “open coding” (Corbin & Strauss, particular core relational themes (Lazarus, 1991) or 1990) individual interviews for perceptions of le- appraisals (e.g., Smith & Ellsworth, 1985). For ex- gitimacy and resistance to change on the basis of ample, when people feel disappointed (unpleasant, “in vivo” descriptions— or those offered by the in- low intensity), they tend to report thinking that the formants—as the basis for establishing first-order situation is unpleasant, that results are below ex- codes (see Van Maanen, 1979). Legitimacy judg- pectations, and that they are certain about what is ments of the change agents are manifest in what happening (e.g., outcomes). When people report respondents said (expressing their thoughts and perceiving the situation as harmful, they perceive feelings) when they evaluated TMs’ actions and having some control over others who intend to non-actions; resistance to change is coded in terms cause harm, anger (unpleasant, high intensity) of what respondents said (expressing their thoughts could be inferred (Roseman, 1991; Tiedens & Lin- and feelings) and what they did (or did not do) in ton, 2001). Ultimately, though, we aggregated these regards to acceptance or support of the change emotional reactions into higher-order constructs, 2014 Huy, Corley, and Kraatz 1659 positive and negative emotional reactions, which judged TMs as legitimate change agents. Our cod- are sufficient to explain the key elements of our ing revealed, for each phase of PROC (formulation, inductive model. implementation, evaluation), shifts in MMs’ legiti- With a rich set of emergent themes in hand, we macy judgments that coincided with (a) shifts in went back to our initial narratives and began map- the emotional reactions MMs had toward TMs ping those themes on to the dynamics we had un- and the change program and (b) a growing resis- covered in the TM and MM descriptions of the tance to the change that culminated in open defi- change effort (Langley, 1999). This mapping af- ance of the TMs’ directives. forded us the opportunity to compare dynamics Figure 2 represents a summary of how the key such as legitimacy judgments, emotional reactions, aspects of our emergent data story interrelated dur- and resistance to change behaviors across different ing each phase, and thus provides a structure for time periods or stages of planned change, as well as our telling the story of Tekco’s PROC effort. For contextual factors (e.g., cultural norms, staff each phase, we list TMs’ attributes and/or actions changes, unexpected internal events, external po- and change outcomes (descriptive aspects of our litical and economic developments) surrounding case), along with MMs’ judgments and emotional these dynamics. Ultimately, this allowed us to be- reactions to TMs and the resulting shifts in their gin establishing the boundaries of a longitudinal acceptance/resistance to the change (from our induc- model for how legitimacy of change agents, emo- tive analysis). We begin with a brief, albeit important, tional reactions, and resistance to change interre- examination of the formulation phase to establish the lated over time. Figure 1 represents our emergent high level of legitimacy and low levels of resistance data structure of MM responses to the change effort, that existed early in the PROC effort. We then spend illustrating the first-order concepts, second-order considerably more time examining the interrelated themes, and aggregate dimensions that serve as the dynamics found in the implementation phase, and foundation for our theorizing (Gioia, Corley, & conclude our data story with the dissolution of the Hamilton, 2013). In this way, Figure 1 provides a PROC effort in the evaluation phase. structured illustration of the links between our raw data and the emergent theorizing that forms the Change Formulation: Favorable Legitimacy cornerstone of our theoretical contribution. The fol- Judgments and Low Resistance lowing sections detail those themes and dimen- sions, and begin tying them together into a coherent Our informants marked the beginning of the understanding of how judgments of change agent change formulation phase with Maxwell’s pro- legitimacy judgments, emotional reactions, and re- nouncement of the need for a radical change. Max- sistance to change interrelate in PROC. well and his new team not only had to sell the need for change, but also the need for change quickly, due to fast-declining market share (from close to FINDINGS 90% only three years earlier down to 75% and As Tekco’s change effort unfolded, it became in- dropping) and profits (a projected decline of approx- creasingly evident to us that the initial acceptance imately 70% the following year). As one executive of the change effort and positive regard the new explained, “The financial markets were clamoring for leaders enjoyed in the first months of their tenure blood.” As a result, time scarcity underpinned MMs’ (during the change formulation phase) would not felt urgency to improve the company’s financial sit- last. Near the end of Tekco’s self-declared imple- uation and reduce its cost structure, as well as the mentation phase, our data provided such a clear decision to forego lengthy detailed strategic planning picture of growing resistance to the change program to achieve major restructuring quickly in three years. that a single empirical question stood out for us: Table 2 provides data examples for this phase (as a What happened to alter the initial acceptance of way to save space for the theoretically more interest- radical change into resistance to change among the ing implementation phase). MMs in this organization? We attempted to empir- Legitimacy judgments: Dominantly cognitive. ically answer this question by tracking the status of As the quotes in Table 2 illustrate, MMs assessed three emergent dimensions (legitimacy judgments, their new leaders as desirable, legitimate agents of emotional reactions, and resistance to change) over PROC dominantly on the basis that new TMs be- the course of Tekco’s PROC effort. As our data longed to a generic category of people who are revealed, MMs evolved in the extent to which they presumed to be appropriate for leading radical  FIGURE 1 Data Structure of Change Recipient Responses to Change Effort 2014 Huy, Corley, and Kraatz 1661 FIGURE 2 Interactions between TMs and MMs through Three Phases of Planned Radical Change change. This category can be described by several company’s top 27 executives, in effect establishing attributes: as newcomers, they were not tied to the a new symbolic basis of authority within Tekco. old regime and they had experience in relevant In addition to MMs forming cognitive legitimacy industry sectors. These attributes were seen by judgments, they also actively interpreted the in- MMs as conforming to cultural expectations set by strumental and relational legitimacy of specific ac- external consultants and industry analysts, thus tions taken by TMs. Newcomer TMs also performed suggesting cognitive legitimacy judgment. Many several additional symbolic actions (Pfeffer, 1981; veteran MMs noted disruptive changes occurring in Zott & Huy, 2007) that reinforced MMs’ prior pos- the broader environment and conceded that radical itive judgment. These actions included involving change was unavoidable, and that it was more ap- MMs in defining the content of the CT program propriate for outsiders to lead radical change than changes (which can be interpreted as enhancing incumbent executives. The “old guard” suffered a TMs’ relational legitimacy). Additionally, the top severe and precipitous loss in status while outsid- team was seen as effectively handling the first ers, in contrast, were taken for granted to be desir- planned layoff in the long history of Tekco, a pro- able leaders of radical change and were presumed posed downsizing of 13,000 positions that repre- to have skills to fight competition successfully. sented around 25% of the workforce. It was feared MMs’ cognitive legitimacy judgment was also the consequences would be devastating to employ- aligned with the thinking of Maxwell himself; in ees who had what amounted to lifetime job secu- his first year alone, he replaced about half of the rity. Yet, this proposed change won the acceptance  TABLE 2 Data Examples for Formulation Phase Legitimacy Judgments Emotional Reactions Resistance to Change Cognitive judgment “I like very much that top managers have given “People simply trust the [new] CEO that he’s “Most management gurus have affirmed that the majority me the opportunity to show what I am capable of, going to do the right thing and will follow of senior managers in protected firms like ours were this is very exciting and I was recognized by my him.” unable to manage in the new competitive context and colleagues. I really liked the fact that I am valued “We believe we do have excellent change should be replaced... Many veteran executives will be by the company, by my bosses, so I am very initiatives here, they are the right solutions replaced, but this is OK.” excited and motivated to drive this change” [also: to improve our operations. We have “[Prior executives’] years of experience were all under implicit positive high activation emotion inferred developed them and believe in them. These one set of rules with one set of orientation and, frankly, through appraisal of personal welfare, benefits changes are going to make a difference [to a very narrow view of the world. So it was rather [new] from being valued by superiors] the quality of our service and reduce our experience that we went to buy in the marketplace.” “The new CEO has done himself some really good cost]. We want to make these changes “If you believe the experts on corporate transformation, service by getting out publicly, showing people happen and see them through.” most would affirm that incumbent executives would be who he is as president of Tekco... He also visited “I would have thought that it would be unlikely to make the transition, so the only viable some locations and met line employees... People harder to bring people along. But, to my solution is to replace them.” liked him and trust that he’s going to do the right great surprise, I see people go along with “I believe that experts who can teach us how to become thing” [also: implicit positive emotion inferred change. I thought that momentum would be more competitive do not come from the inside.” through appraisal of potential benefit for the firm slow at the beginning, one had to show “We wanted to change the old Tekco monopolistic with a likable and trustworthy CEO] positive results for momentum to slowly culture... Most of the new executives who came in “I think many of us here were selected for our pick up.” did not come from monopolies, so that’s good. Most of ability... our interest in making changes, and “We are extremely convinced about the need them came from free market, free enterprise industries achieving a challenge and being very goal for a comprehensive transformation of the like GE and IBM.” focused... We have seen the importance of company, we understand very well our Relational judgment these initiatives for the future, and we have been [dire] financial situation.” “I truly believe the way Tekco downsized [shows the recognized [by top managers] for doing that. So “We are now facing new market conditions, organization] has been a very good corporate citizen. If that feeds many of us... because we are goal the entry of international competitors, new you look around, their termination packages are as good focused and achievement oriented” [implicit client expectations, so we don’t have much as anybody’s.” positive emotion inferred through appraisal of choice about not changing drastically.” personal benefits from being valued by superiors] “We feel the anxiety and the urgency for “We are getting increasingly confident that we can change in this highly competitive overcome most things that can get thrown at us. environment. We bought entirely into the We have sufficient determination to deal with it” need for change and want to help it [implicit positive or neutral emotion inferred succeed.” through appraisal of no personal harm, and even some optimistic confidence in taking risk and overcoming challenges] “I truly believe the way Tekco downsized [giving competitive benefits] to be a real tribute to a caring organization” [implicit positive or neutral emotion inferred through appraisal of caring or no harm] 2014 Huy, Corley, and Kraatz 1663 and support of MMs, thanks, in large part, to pro- [new CEO] has done a very good job as our leader mulgated symbolic principles by the CEO and his and external ambassador. He’s doing extremely top team (again, see Table 2 for data examples). In well and employees in [Tekco] look to him as being addition to offering relatively generous compensa- their leader.” tion to the victims of the downsizing, Maxwell also Despite these positive judgments and actions decreed: “We are going to remove the work before from MMs, two critical commitments were made we remove the job.” This symbolic pronouncement during this formulation phase that would prove also helped generate support among MMs, who problematic in later phases: (1) Maxwell’s very vis- were involved in designing their own cost-cutting ible pledge to remove work before removing jobs projects based on promised new IT and believed and (2) TMs’ promise to deliver to the board of that the CEO’s promise that work would be re- directors (and shareholders) $700 million annual moved before people would not materially reduce net income at the end of a three-year change effort. the quality of customer service, a key component of When asked why the three-year implementation Tekco’s identity to this point (which could be in- period was chosen, CEO Maxwell explained: terpreted as enhancing instrumental legitimacy). Two years is too fast to go from implement and An HR survey of employees reported that many achieve; five years is too long. You had to be pres- affected employees viewed the downsizing rules to sured by time schedule. You could only put in the be “fair” and “kind,” even though the implicit psy- window the period of time you could sell to your chological contract of lifetime job security had been shareholders. So everybody accepted the three-year violated (relational legitimacy). TMs’ additional plan, and that becomes the base around which we symbolic actions thus conveyed various types of measure ourselves to the outside world and to the legitimacy in addition to the dominant cognitive shareholders and investment community. legitimacy. Emotional reactions. As could be expected in Change Implementation: Shifting Legitimacy the wake of positive judgments, MMs’ emotional Judgments and Increasing Resistance reactions toward TMs were largely positive to neu- tral. More specifically, for MMs who championed Ironically, many of the same factors that pro- change projects, collective emotions were largely duced high legitimacy and low resistance during positive, including such emotions as excitement to the formulation phase came back to haunt execu- realize new projects that they proposed and hope tive change agents in the implementation phase. for a better future for their professional develop- Our analysis suggests that emerging and protracted ment and for the improved performance of the com- challenges related to the implementation of various pany. Other MMs—in particular, those in the oper- radical change projects led MMs to reassess their ating line groups who would implement the original legitimacy judgments in the face of mount- changes proposed by their colleagues while trying ing implementation problems and heightened neg- to maintain the quality of customer service— dis- ative emotional reactions. Over the course of the played calm (positive), resignation, and neutral implementation phase, these negative emotions emotions (neither positive nor negative) toward and modified legitimacy judgments fueled a grow- TMs. Overall, MMs’ emotional reactions toward ing resistance to the change that eventually led to TMs were positive or neutral, with very little neg- the downfall of the PROC program. ative emotions toward TMs or the PROC effort. Interim change outcomes. Despite positive feed- Actions. Contrary to intuitive expectations and back from external analysts and an increase in research about MMs resisting major change due to stock price from $28 to $39 in less than a year, self-interest (e.g., Biggart, 1977; Tichy & Sherman, employees began to see cracks in Maxwell’s change 1994), there was little discernible resistance to plans: a random telephone survey of about 1,800 change, even from MMs who did not belong to the employees showed that only 38% of employees felt change formulation groups. A company study that the company had a sincere interest in their showed that more than 80% of the surveyed em- wellbeing, 61% rated their morale from fair to very ployees agreed that Tekco’s financial performance poor, and only 34% believed that customer focus had to improve if it was to survive in a competitive would be better off at the end of the three-year environment. A veteran senior MM in HR summa- change effort. Underlying this internal concern was rized the results of her company’s employee survey the realization among MMs that many CT projects near the end of the formulation phase thus: “The were meeting delays or producing lower than ex- 1664 Academy of Management Journal December pected benefits. Recall that the design of the change interim results in mid-course, to re-prioritize re- projects was done very quickly and superficially sources, or to re-adjust plans was unwelcome. As during the formulation phase. MMs now blamed one manager, analyzing the evolution of the port- TMs for not giving them adequate face time and folio of change projects, said: support to resolve these difficulties effectively and It’s like 155 lottery tickets all winning on December quickly. 31, 2006... There’s now a huge organizational As one operations MM championing a number of resistance to [re-adjustment]. We don’t like to go change projects reported: back and say we made a mistake. We’re going to My [senior executive] stayed at a very high level, waste valuable company time, dollars, and skills. wanted to manage the numbers, to understand some People are dedicated to just doing their own proj- of the solutions, but really did not provide time to ects, and they worry about whether their initiative do that. It was very difficult... He would always tell will get [enough resources to be] implemented, as us that he’s available anytime, but he was so busy opposed to [asking themselves], “Have I got the right that you’d get a half an hour here, and, two weeks initiative?” later, you’d get an hour there... We sensed that he This lack of engagement with MMs was also rec- really didn’t understand, not because he wasn’t ca- pable of understanding, but that he wasn’t focusing ognized by TMs. After having initiated CT and on some of the more detailed [aspects]. putting a control structure in place, CEO Maxwell devoted most of his time on external stakeholders. Another operations MM added: He described his focus in a private interview as We did not succeed in getting our bosses to go out to follows: rally our employees. Most of their communication was written, formal communication. We were think- I expect that [people below] will deliver. My role ing of a training package to help them improve face- now becomes one of managing from altitude the to-face communication. work, explaining to the public what Tekco is all about and where it’s going. Explaining to the gov- Rapid launch of change projects that generated ernment, talking to investors, and trying to keep MMs’ favorable judgment of TMs as competent myself up to date with where the industry is going. change agents during the formulation phase thus That keeps me fairly well employed. began to hinder the smooth implementation of var- While Maxwell assumed that other senior exec- ious change projects, foreshadowing real prospects utives would spend much of their time champion- of underperformance and threatening MMs’ sense ing and selling the change efforts inside Tekco, this of self-worth. MMs needed to regularly justify to did not take place. These executives reported in other colleagues that their championed projects re- private interviews that they were also over- mained economically and operationally viable, and whelmed with tasks similar to those of the CEO. they blamed TMs for having created this undesir- To make matters worse, the generous downsizing able time scarcity and performance pressure envi- compensation terms planned in the formulation ronment. Rapid launch in the formulation phase— phase also generated unintended effects on the op- desirable in the eyes of both TMs and MMs—now erations groups: 16,000 employees opted to leave, became a source of disappointment and anxiety. 3,000 more than planned. Remaining employees “Failure in not an option” (a line from the movie felt overly burdened, especially supervising line Apollo 13) was often repeated in MMs’ project re- MMs who were responsible for customer service. view meetings. As one MM in the control group Even though top executives were not directly re- observed: sponsible for this unexpected event, MMs still held When these targets were set, there was nothing left them responsible as heads of the company and for we could do except to try to make things work, to forcing them to cut employees and costs to meet the deliver these projects while we were locked in some financial targets as per the initial time schedule. kind of a time box, with unrealistic schedules and TMs’ promise “not to remove people before work” [financial] benefits. This immediately created much had not been respected. All this caused a growing tension between the [recipient line] groups who change fatigue among employees from having to must endorse these benefits and [the change agent work so hard, so fast in such a compressed time groups who proposed these benefits]. frame. As one customer service representative la- Project MMs tried desperately to realize their mented on behalf of her colleagues: “There’s so change projects, and any suggestion to evaluate the much information that we’ve become saturated. 2014 Huy, Corley, and Kraatz 1665 We’ve become so saturated that we’ve become stu- with customer service—felt disappointed by TMs pid. So very simple things become difficult.” A not keeping their promise of not removing employ- senior project manager in charge of many projects ees before work was first streamlined. These line that required major IT development also noted the operations MMs were under intense pressure to effect this fatigue had on the implementation effort: deliver good customer service while managing the collective anxiety and resentment of their line em- There is a lot of fatigue, the problems are enormous, ployees who were learning new competencies I don’t know many people who work less than 70 to 80 hours a week... People in [IT] Systems Devel- while at the same time subject to downsizing. For opment work very hard, for long hours, and many of instance, a business unit composed of about 3,500 them fell sick physically and mentally. Many de- customer service representatives had to go through pleted their energy because they could not sustain a major downsizing and geographical consolidation such an intense level for such a long time. effort. One MM in charge of about 800 of those representatives reported: Some MMs in HR and operations knew the grav- ity of such actions and had been pleading with the Our customer representatives are getting tired from top team to issue a public and honest explanation, all this long traveling [due to centralization of cus- but this did not take place (moral judgment). tomer service centers in major metropolitan areas] Emotional reactions: Shifting to negative. and this situation will not be sustainable in the long These mounting implementation problems resulted run. Many are mothers with young children. As a in MM emotional reactions shifting from positive/ result, sales revenues are going down: senior man- neutral to negative emotions such as disappoint- agement asked me to put the screws on them even more. I feel stuck... I have to sell this change to my ment, anxiety, and frustration. There were subtle employees but they don’t believe me. differences in MMs’ emotional reactions depending on the dominant role they performed, however. As one line MM described her employees’ emo- MMs as change project champions felt disap- tional reactions: pointed because they were not getting adequate interaction and support from TMs (relational judg- Employees told us to stop saying that change would be faster, we were making their lives more difficult ment). MMs attributed their mounting difficulties by moving them around, yet they did exactly the in implementing their change projects to insuffi- same job as before, they answered the customers the cient TMs’ support. This anticipation of underper- same way as before, so nothing important in their formance, which could have negative identity and work has changed... They only see that they have career consequences, also led to anxiety in dealing less resources, less time to devote to customers, less with TMs. One MM described the emotional reac- time for training... We had tremendous difficulty tions of the change group as follows: in delivering good service this year. People are so scared not to deliver. They are no These employees’ emotional reactions clearly longer asking the real questions, not only because had an emotional effect on their supervising MMs. they don’t know what they are, they’d rather not These MMs, who were emotionally neutral and think about them... we are playing a little bit the accepted the need for radical change during the ostriches... people will try to deliver at all costs. change formulation phase, became increasingly In aggregate, there was mounting shared frustra- critical of both TMs and other MMs who promoted tion that their company was going in circles and radical change. Their emotional reactions could be wasting resources in the change process, especially best described as resentment. As one MM said, in the wake of the top team’s inability to articulate “They say that the greatest priority is our custom- a meaningful change goal for the organization after ers. Yet all the actions demonstrate that the highest the first year’s learning grace period had elapsed. priority is short-term financial results... People As an MM in operations voiced in frustration: are not that naïve.” Collective anxiety also prevailed among these There’s a lack of understanding [about] what the recipient line MMs, as one MM made very clear: future is going to hold, and how we’re going to deal “Everybody right now is feeling highly insecure, with it. There’s a big void... People are frustrated. they are wondering if they would continue to have Like, are we just doomed? a job in a year from now”; and, again: “This creates Meanwhile, MMs in the role of pure change re- a very high level of stress. Some of us are stressed, cipients—typically, in operations groups charged de-energized, fed up... We are in hell.” Another 1666 Academy of Management Journal December senior HR MM validated this widely shared emo- attributions were so prevalent that even the CEO tional reaction of MMs toward TMs: “I’m disap- became aware of them: he mentioned these phrases pointed on a lot of counts. I’m disappointed be- in a private interview. These attributions marked a cause the leadership of this company, at the very shift in MM legitimacy judgments, from the domi- senior levels, has not delivered on its responsibili- nantly cognitive base of the formulation phase to ties.” These MMs’ negative judgments and emo- more of a concern with the relational and moral tions were validated by the CEO himself in a pri- aspects of legitimacy. vate interview toward the end of the second year Relational legitimacy. Judgments involving how of CT: a social entity communicates with others in a way that accords them respect, dignity, and status Over the last three months, I went out and began within their social milieu constitute the basis of touching some of the people and I didn’t like what I relational legitimacy (Tyler & Lind, 1992). If people was hearing. It showed me that the senior executives perceive that their self-worth and identity are en- in this company didn’t stay in touch with the peo- hanced in this interaction, they will perceive rela- ple. I didn’t stay focused enough with the senior tional legitimacy to be high (Tost, 2011). MMs who executives... We left the employees on their own. were co-opted by top executives to design the con- I think that’s a huge mistake. tent of radical change felt well supported by their Emotional censorship. MMs’ moderately nega- superiors during the formulation phase and thus tive emotional reactions were exacerbated by the had assumed this high level of support to continue top team’s deliberate and public banning of any during the implementation phase. But top execu- expression of doubt and cynicism by MMs (as those tives’ lack of engagement and implementation mis- negative emotions were deemed too high a risk for cues reduced MMs’ sense of support and gradually demoralizing lower-level workers). An executive impacted the sense of respect and status felt ac- memo was circulated to MMs decreeing that “Cyn- corded to them by senior change agents. Disap- icism will not be tolerated; we are in a position of pointment with change project progress and per- leadership and must project hope and confidence ceived insufficient TMs’ support also led some in difficult times.” This formal articulation of a MMs to make harsh attributions of the top team’s norm that was already intuitively known by many motives. One MM shared this negative appraisal: MMs puzzled them, but effectively served to shut [Executives were] “jerking them around all the off any legitimate outlet for expressing negative time,” and that, “they are here just for the short emotions to the top team. At the lower-worker lev- term... [they] appear ready to do anything to els, the feeling of emotional censorship was less achieve short term results.” Relational and moral acute because many MMs organized venting ses- judgments can blend: If people feel they are treated sions. However, such sessions were considered with disrespect, they can make an additional judg- risky. Those who had managed to survive the lay- ment that the disrespectful person is low in moral offs thus far dared not express their true feelings character. Not fully aware that TMs were over- lest they be taken advantage of one day. As one MM whelmed with external constituencies such as gov- in operations described, “You have to show enthu- ernment agencies and major customers, MMs siasm for a new idea even if you don’t believe in it, tended to attribute insufficient two-way, authentic so you have to fake your enthusiasm.” This, in turn, communication either to TMs’ personal reluctance reduced dialogue for understanding between the for honest dialogue or to their under-appreciation top team and MMs. of the importance of communicating about change. Legitimacy judgments: Dominantly relational As one MM from HR conveyed harshly: and moral. The shift to more moderate negative emotional reactions fostered mounting unfavorable [Executives] will tell you over and over again, “You judgments about the top team’s legitimacy to lead can’t communicate enough.” But, here they are, these champions of change, and, for whatever rea- change