HOW MIDDLE MANAGERS’ GROUP-FOCUS EMOTIONS AND … › quy-huy › documents › SMJ 2011... · 2015-11-27 · Middle Managers Group-Focus Emotions and Strategy Implementation 1389

Strategic Management JournalStrat. Mgmt. J., 32: 1387–1410 (2011)

Published online EarlyView in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/smj.961

Received 12 January 2009; Final revision received 2 August 2010

HOW MIDDLE MANAGERS’ GROUP-FOCUSEMOTIONS AND SOCIAL IDENTITIES INFLUENCESTRATEGY IMPLEMENTATION

QUY NGUYEN HUY*Strategic Management, INSEAD, Singapore

The literature on top-down strategy implementation has overlooked social-emotional factors. Theresults of a three-year field study of a large technology firm show how top executives who favoran affect neutral task approach can inadvertently activate middle managers’ organization-relatedsocial identities, such as length of time working for the company (newcomers versus veterans)and language spoken by senior executives (English versus French), generating group-focusemotions. These emotions prompt middle managers—even those elevated to powerful positionsby top executives—to support or covertly dismiss a particular strategic initiative even when theirimmediate personal interests are not directly under threat. This study contributes to the strategyimplementation literature by linking senior executives’ actions and middle managers’ socialidentities, group-focus emotions, and resulting behaviors to strategy implementation outcomes.Copyright 2011 John Wiley & Sons, Ltd.

INTRODUCTION

Although the essential role of middle managers ininfluencing the process and outcomes of strate-gic change seems well established (e.g., Balo-gun and Johnson, 2004; Floyd and Wooldridge,1992; Huy, 2002; Westley, 1990), the reasons mid-dle managers support or oppose strategic changehave often been assumed to be mainly driven bytheir personal interests in the narrow sense of theterm (e.g., gains or losses in their organizationalstatus, role, or economic incentives) (e.g., Big-gart, 1977; Crossan and Berdrow, 2003; Guth andMacMillan, 1986). Research on implementation ofstrategic change has underexplored causes that areless directly personal but could still significantly

Keywords: group-focus emotion; strategy implementa-tion; middle manager; social identity; qualitative research∗ Correspondence to: Quy Nguyen Huy, INSEAD, StrategicManagement, 1 Ayer Rajah Avenue, 138676, Singapore.E-mail: [email protected]

influence middle managers’ support or dismissalof a new strategy.

The fact that individual middle managers expe-rience positive (negative) emotions as a result ofsome perceived beneficial (harmful) impact ontheir narrow personal interests and that those emo-tions influence their willingness to support (dis-miss) a new strategy seems relatively intuitive. Butscholars have yet to investigate middle managers’group-focus emotions, those individual emotionsthese managers feel on behalf of a group or fellowgroup members who experience a specific eventeven when they themselves are not personallyaffected by it (Smith, Seger, and Mackie, 2007).1

For example, middle managers who are indi-vidually elevated by senior executives to occupypowerful positions in strategic change can still feel

1 Although Smith et al. (2007) use the term ‘group-level emo-tion’ to describe this construct, I substitute the term ‘group-focusemotion’ to reduce potential confusion with collective emotionsin the group (or emotions shared among many members of thegroup) (e.g., Barsade, 2002).

Copyright 2011 John Wiley & Sons, Ltd.

1388 Q. N. Huy

angry that other middle managers are mistreatedby these executives and act against them. Thus,group-focus emotions are distinct from collec-tive emotions, which refer to emotions that areshared among many members of the group—oftenthrough social interactions, including emotionalcontagion (Barsade, 2002). Group-focus emotionsare felt privately by individuals; it is not neces-sary for them to be expressed or shared with otherpeople. Yet it is possible for many members ofa group to feel the same group-focus emotionsif, for instance, they interpret an event in a sim-ilar way (Smith and Ellsworth, 1985). Emotionsare important to study because empirical researchhas shown that emotions can significantly influ-ence the quality of thinking and behavior which,in turn, affects performance in organizational set-tings (Barsade, 2002; George, 1995; van Zomeren,Spears, and Fischer, 2004).

Recent experimental research has suggested thatgroup-focus emotions can be elicited by identifi-cation with one’s salient social identities (Smithet al., 2007). But we have insufficient empiricalunderstanding of what specific social identitiesmight be triggered among middle managers in atop-down strategy implementation process, whatspecific group-focus emotions are experienced bymiddle managers, how these actually affect theirthinking and behavior, how top executives dealwith these social-emotional issues, and how all thisaffects the success of top-down strategy implemen-tation. Consequently, I have used inductive qualita-tive research to study a top-down strategic changeoccurring in a large bilingual (French and English)Canadian information technology company in realtime over three years.

According to Suddaby (2006), this groundedresearch method is best suited for research ques-tions that require understanding the process bywhich actors interpret and construct meaning outof intersubjective experience. Pratt (2009) fur-ther noted that qualitative research is best suitedto addressing the ‘how’—as opposed to ‘howmuch’—question and for understanding the worldfrom the perspective of those studied, as well as forexamining and articulating processes. Thus, qual-itative inductive research complements hypothesistesting research by providing a contextually richunderstanding of the links among various con-structs and outcomes or by revealing new con-structs and processes in natural organizationalsettings.

This study generates several findings. First,despite the seemingly intuitive importance ofattending to middle managers’ emotions in imple-menting a top-down strategy, most top execu-tives in this study did not give high priority tothis matter, favoring a nonaffective task focuseven when they encountered repeated difficultiesin implementation. Second, for middle managers,the notion of self-interest encompassed more thanissues of individual power, status, and economicbenefits. Rather, it included organization-relatedsocial identities such as tenure (i.e., length of timeworking for the company, which conveys organi-zational loyalty, experience, and competence thatdistinguish veteran middle managers from new-comer executives) and language (i.e., French ver-sus English spoken by executives). Third, thesesocial identities elicited middle managers’ group-focus emotions and stimulated them to dismiss orsupport a specific strategic initiative, even thoughtheir direct personal interests were not at stake.Finally, this study also shows that top executivescan adjust their strategy implementation actionsand improve the odds of success of strategicchange if they seek to understand and address thecauses of middle managers’ group-focus emotionsin situ.

This research contributes several insights to theliterature on top-down strategy implementation inregard to the nature of middle managers’ emotions,their social identities at work, and how they canbe motivated for strategic change. As noted, thisstudy reveals causes that are less directly personalbut could still influence middle managers’ supportor dismissal of a new strategy. The study com-plements relatively decontextualized model test-ing research with a culturally grounded, proces-sual analysis of how group-focus emotions ariseover time and influence the success of strategyimplementation.

Focusing on middle managers’ group-focusemotions in top-down strategy implementationextends our understanding of various causes oftheir support or opposition in several ways. First,scholars of strategy implementation have tendedto adopt an affect neutral perspective that focuseson bounded rationality, cognition, and task-relateddimensions. For example, implementing a newstrategy can be more successful if top executivesexpress their vision clearly and design appropri-ate task systems, such as formal structures, controlsystems, and incentives. The business logic of the

Copyright 2011 John Wiley & Sons, Ltd. Strat. Mgmt. J., 32: 1387–1410 (2011)DOI: 10.1002/smj

Middle Managers Group-Focus Emotions and Strategy Implementation 1389

new strategy and the logic of congruence amongvarious task systems are assumed to align behav-ior at all levels of the organization (e.g., Bour-geois and Brodwin, 1984; Daft and Macintosh,1984; Govindarajan, 1988; Nutt, 1989). Further-more, researchers who have adopted a cognitivefocus have explored how organization membersmake sense of strategic events and how their inter-pretation influences later actions (e.g., Balogun andJohnson, 2004; Gioia and Thomas, 1996; John-son, 1988). These research perspectives, althoughinsightful, have overlooked how middle managers’emotions can help or hinder top-down strategyimplementation. Because middle managers occupyleadership positions, their emotions can influencemany of their subordinates’ emotions, thinking,and behavior (Sy, Cote, and Saavedra, 2005), allof which can influence strategy implementation.2

In a similar vein, emotion has often been con-sidered an individual phenomenon, bearing onpersonal concerns such as an individual’s owngoals or resources. For example, appraisal the-ories of emotion (Ellsworth and Scherer, 2003;Frijda, 1986; Smith and Ellsworth, 1985) viewemotions as arising when a person appraises anevent as harming or helping her goals. As Smithet al. (2007) noted, this approach does not fullyexplain the kinds of group-focus emotions that areelicited when, for example, people are elated whentheir sports team wins or saddened by casualtiescaused by natural disasters that do not directlyaffect themselves or their families.

Finally, this research contributes to the man-agement of strategic change in multicultural firmsin which employees with diverse social identi-ties, including nationalities and languages, workwith one other. When firms use means such asmergers and acquisitions, strategic alliances, andjoint ventures with foreign counterparts to increasetheir competitive advantage (Gomez-Mejia andPalich 1997; Rangan and Sengul, 2009), they facethe challenge of managing strategic change witha diverse workforce in which employees exhibitdiverse social identities such as nationalities andprofessional identities (e.g., biomedical and engi-neering). Much empirical research on strategic

2 The terms ‘strategy implementation’ and ‘strategic change’ willbe used interchangeably in this text. This study describes atop-down strategy implementation, which constitutes a specialform of strategic change. The process of strategic change canhave many forms, such as top-down directed, bottom-up andemergent, or combining aspects of both.

change seems to have assumed or occurred in arelatively homogeneous workforce (e.g., Balogunand Johnson, 2004; Fiss and Zajac, 2006; Gioiaand Thomas, 1996; Pettigrew, 1985). We know,therefore, relatively little about how a diverseworkforce responds to top-down strategic changeand how these responses influence the outcome ofsuch change.

In particular, although we know that strategicchange can arouse strong emotions among theworkforce (emotions that can influence employ-ees’ receptivity to change (e.g., Fox and Amichai-Hamburger, 2001; Huy, 2002)), empirical researchhas underexamined how employees’ emotions in adiverse workforce influence the outcomes of strate-gic change. Several scholars (e.g., Vaara et al.,2005) have started to explore how emotion-ladenlinguistic differences influence cultural identitybuilding in a merged firm, but we don’t know suffi-ciently how these differences influence the successof strategic change.

In this paper, I first review the literature toarticulate some of the potential conceptual linksbetween group-focus emotions and social identi-ties. Second, I describe how I collected and ana-lyzed the qualitative data. Third, I present vari-ous cases of top-down strategy implementation inwhich I identify the causes and consequences ofmiddle managers’ group-focus emotions and howtop executives dealt with them. I end with implica-tions for managerial practice and future research.

Group-focus emotions, social identities,and strategy implementation

Because the findings suggest some organization-related social identities as the main cause ofgroup-focus emotions that influence the success oftop-down strategy implementation, a review of theliterature on group-focus emotions and social iden-tity can help us better understand the data analysisand assess the plausibility of the findings.

Experimental studies involving students haveshown that group-focus emotions predict collec-tive behavior (e.g., ingroup support and outgroupconfrontation) much more strongly than other indi-vidual emotions do (Smith et al., 2007). This likelyhappens because perceived consensus within peo-ple’s important groups can help define realityfor them (Hardin and Higgins, 1996). In con-trast, a person who experiences more idiosyn-cratic, personal emotions—arising, for example,


1390 Q. N. Huy

from interpretations that she suspects may be dif-ferent from those of other group members—mayharbor self-doubt about the social validity andappropriateness of her private feelings and mayfeel less confident in expressing and acting onthose feelings, especially in ways that might beperceived by other members as violating groupnorms (Hochschild, 1983). This person could pri-vately experience emotional dissonance and stress(Grandey, 2003), but the effect on group-leveloutcomes is less clear. Just as belief consensusincreases certainty, reduces anxiety, and motivatesaction (Milgram, 1992; Turner, 1991), interpretingthat others likely share the same emotions relatedto the same causes increases one’s propensity foraction.

Scholars have shown through experiments thatmere consideration of a group one strongly identi-fies with can elicit group-focus emotions, such aswhen an individual thinks of himself as a U.S. cit-izen or as a Republican. Moreover, members whoidentify strongly with their group likely experienceemotions similar to those of others in the groupwhen faced with events that affect the group’sidentity or welfare (Dutton and Dukerich, 1991).

The social identity literature helps us understandsome of the psychological mechanisms that under-pin group identification behaviors. Social identityrefers to ‘the socially shared understanding of whatit means to be a group member, and this typi-cally includes stereotypes of in- and out-groupsin relation to each other as well as apprecia-tions of the relative status of those groups’ (vanZomeren, Postmes, and Spears, 2008: 505). Socialidentities are important to people’s self-image andself-esteem, and events that make these identi-ties salient can invoke strong emotions (Kreiner,Ashforth, and Sluss, 2006). People typically seethemselves as members of a number of discretecategories (e.g., nationality and profession). Eachof these memberships is represented in an individ-ual’s mind as a social identity that both describesand prescribes what he should think and feel andhow he should behave (Tajfel and Turner, 1979).

Group social identities are shaped by at leasttwo psychological processes: self-categorizationand self-enhancement. Self-categorization mini-mizes ingroup differences and maximizes inter-group differences (Turner et al., 1987). Individualsbecome group members in that they think and actas embodiments of the relevant ingroup prototyperather than as unique individuals. A prototype is a

subjective representation of the defining attributes(e.g., beliefs, attitudes, and behavior) of a socialcategory, constructed from relevant social infor-mation, particularly information that emphasizescontrasts with outgroups. Social identity is, thus,dynamic and contextual as it responds to immedi-ate social comparisons with various outgroups thatbecome salient in particular situations (Ashforth,Harrison, and Corley, 2008).

Self-enhancement is the second underlying pro-cess central to shaping social identity. It assumesthat people have a basic need to achieve and main-tain a positive sense of self. Group members arestrongly motivated to engage in collective self-enhancement by making comparisons on selectivedimensions that favor the ingroup and disadvan-tage the outgroup (Hogg, Terry, and White, 1995).Valuing group social identities can elicit positiveemotions such as joy and pride; devaluing iden-tities can arouse negative emotions such as angerand fear (Mackie, Devos, and Smith, 2000). Bothof these aspects of group social identity—self-categorization and self-enhancement—enable peo-ple to transform their emotional experience of par-ticular events into a more powerful collective force(Haslam and Reicher, 2006).

Although scholars have attracted our attention tothe importance of organizational identity in changecontexts (e.g., Corley and Gioia, 2004; Fiol, 2002),we know less about how other types of social iden-tity influence top-down strategy implementation.Because each employee can have many differ-ent identities (e.g., gender, profession, age, workgroup, nationality, religion), it can be difficult forleaders of strategy implementation to know aheadof time which identity matters most to employ-ees in a given situation. Because emotions arearoused only about issues people consider impor-tant to them in a given situation (Lazarus, 1993),the identities that matter most are likely to be theones that are associated with emotions, positiveor negative. Thus, emotions represent a situatedmarker of identity salience, and group-focus emo-tion represents a situated marker of group identitysalience in a particular situation.

METHODS

Research setting

Canco (a pseudonym) is a large Canadian informa-tion technology service company. Founded more



than 60 years ago as a small venture, Canco hadmore than 50,000 full-time employees, net assetsof about $10 billion, and a market value exceed-ing $12 billion at the time of this research. Becauseof its dominant market position, the company hadbecome a quasi monopoly subject to governmentregulation.3

Then came industry disruption. Deregulation inlate-1992, followed by the entry of competitors,changed the market rules. A radical and rapidchange in strategy and structure was required toaddress the shift from weak national competi-tion to strong global competition. A vicious pricewar ensued. Annual profits declined by 40 per-cent in the first year of competition, and the com-pany faced rapid market erosion of about 12 per-cent annually. Sweeping changes were needed.A newcomer, Richard Foster4, was appointed aschief executive officer (CEO). Within a year, hehad decided to radically change Canco’s corpo-rate strategy through a series of strategic initiativesthat altered the firm’s formal structure, incentiveand control systems, competencies, personnel, andculture. These initiatives shaped what was knownwithin Canco as the corporate transformation (CT)program, which was to unfold over three years.

Having had good working relations with manymiddle managers and executives in various depart-ments of Canco over the previous 10 years, I askedfor permission to interview them privately and reg-ularly for the next three years (mid-1995 through1998) and observe company meetings. I explainedto them that real-time longitudinal research onstrategic change is rare and that my impartialresearch could bring valuable learning to them. Toensure minimal interference and protect the confi-dentiality and anonymity of all respondents, it wasagreed with both management and unions that Iwould not give interim feedback to anyone dur-ing the four years of data collection and analysis.As I demonstrated discretion and trust built up,they allowed me to tape interviews and accessconfidential document records. Interviews lasted,on average, from 60 to 90 minutes. I obtained anemployee pass that allowed me to circulate freely

3 Much company-specific information was made vague to dis-guise the company, which is still operating. Numbers may varyby plus or minus 20 percent. These modifications should notinfluence the conceptual findings discussed in this paper. Cur-rency amounts are in U.S. dollars.4 All individual and company names are fictitious to protectanonymity.

on Canco premises and talk to whomever I wished.On average, I spent one to two days a week forthree years at Canco to observe meetings and con-duct interviews. I spoke both French and English,so I interacted with employees in the language oftheir choice.

My initial research question was open ended:How does the evolution of employees’ percep-tions, emotions, and actions affect the outcomeof major change projects? During regular inter-views, I asked general questions such as: Canyou describe various change projects you wereinvolved with, either as change recipient, changeagent, or both? What actions did you take andwhy? Please describe the interactions you have hadwith various groups. What reactions did you expe-rience? Please describe your thoughts and feelingsin relation to events you believe are significant toyou. What did you do as a result of these thoughtsand feelings? What outcomes or business resultsdid you achieve? Please name other people withwhom you have interacted who might give me aricher or different perspective.

The idea of building a conceptual model basedon qualitative data on interactions among peoplein a single company was inspired by the work ofscholars such as Bartunek (1984) and Gioia andChittipeddi (1991). As explained earlier, focusingthe current analysis on the emergence and con-sequence of middle managers’ group-focus emo-tions and social identities came later, during itera-tive data analyses and thanks to recent conceptualadvances in the emotion literature. I became awareof the concept of group-focus emotions (Smithet al., 2007) 10 years after the end of my data col-lection.

Data collection

I conducted more than 200 informal conversa-tions with about 200 employees at all levels ofthe company (most of whom I had already hadsome interactions with over the previous 10 years)to corroborate and triangulate what I had seen andheard and to watch for various kinds of change.This process allowed me to screen the people mostlikely to provide insightful and differentiated per-spectives. I then asked them for regular formalinterviews, during which I often asked them tosuggest at least two other employees they consid-ered influential in the realization of one or several


1392 Q. N. Huy

projects, particularly those who might not con-cur with their own perspectives. I repeated theprocedure until various perspectives were cited atleast twice (Laumann and Pappi, 1976). To explorethe research question, I interviewed the incumbentCEO three times, the two previous CEOs onceeach, 12 executives 26 times, and 114 middle man-agers 192 times over three years, in addition tolower-level employees and union officials.5 Morethan 75 percent of these interviews were taped andtranscribed. I had an office on-site, and I had manyopportunities to observe people at work, interactwith them, and observe meetings. The companyalso provided numerous reports on change projectstatus and employee surveys. Several secondarydata collection techniques, including observationsand perusal of archival data, allowed me to triangu-late findings from diverse sources to build strongerassertions about interpretations (Eisenhardt, 1989;Yin, 1994).

Data analysis

Preliminary analyses suggested that managersexperienced a wide range of emotions (positiveand negative) related to various change initiatives,whether they led those initiatives or were on thereceiving end. As noted earlier, the fact that theyfelt emotions because of events that affected themdirectly (e.g., their own work effectiveness or per-sonal status in the company) seems relatively intu-itive. What is less intuitive, however, is that therewere a number of middle managers who were pro-moted or reappointed by the new top executiveteam to powerful positions (e.g., vice presidentor director) who still experienced emotions aboutevents that did not affect them personally (whichI refer to as group-focus emotions). What is alsointriguing is that top executives were often not ableto identify these group-focus emotions and theircauses linked to social identities, let alone addressthem.

5 Middle managers are two levels below the CEO and one levelabove first-line supervisors. There are many levels of middlemanagement at Canco, so one senior middle manager in the linegroups could be in charge of 2,000 to 5,000 front-line workers(Uyterhoeven, 1989). Such a manager has authority over juniormiddle managers; they, in turn, have a typical supervisory rangeof 50 to 100 workers. At Canco, people with titles such as vicepresident, director, and associate director are treated as middlemanagers. Top management includes the CEO, and those whoreport directly to him bear the titles of executive or senior vicepresident or group vice president. Senior management includespeople at these upper levels.

As the number of middle managers I interviewedwas large (114 middle managers in various groups,including general management, human resources,finance, operations, and sales and marketing), Ifocused my analysis on a more manageable subsetof accounts of middle managers from two func-tional groups: human resources (HR) and sales andmarketing (SM). Several reasons motivated thischoice. Middle managers in the HR group workedas linchpins across various groups in Canco andacross all hierarchical levels. Because of theirroles, HR middle managers continuously collectedand had access to private information about thepersonal profile, task performance, leadership abil-ity, and morale of each employee and group inCanco. HR middle managers also continually con-ducted private interviews and group interventionsacross the company, so they could report on pat-terns they observed across various groups. As aresult, I analyzed 27 private interviews with 15HR middle managers.

As one successful initiative to be reported latertook place in the SM group, I also analyzed allof my 42 interviews with 22 middle managers inSM. I analyzed the perspectives of members of thetop management team by analyzing the accountsof the CEO, whom I interviewed privately threetimes during the three-year corporate transforma-tion process. I also analyzed eight interviews con-ducted with the top executive of the SM group andfour senior executives working closely with the topteam (two in corporate strategy and two in SM). AsEisenhardt and Graebner (2007) suggested, inter-viewing informants from diverse groups and per-spectives reduced the risk that all these informantswould engage in biased, convergent retrospectivesensemaking and/or impression management. Thequasi-real-time data collection used in this studyalso mitigated rationalization biases. In sum, I ana-lyzed 69 interviews with 38 middle managers and11 interviews with six members of the senior exec-utive team.

These middle managers formed a diverse groupin terms of tenure (veteran/newcomer) and gender.All of them, both in HR and in SM, shared oneimportant characteristic: they were all appointedor reappointed by the new top executive team tooccupy influential positions at Canco and, as val-idated in private interviews, did not feel direct,immediate threat to their personal status and privi-leges under the new regime. But some of them did



feel group-focus emotions linked to social identi-ties, as will be discussed.

To code emotions in respondents’ accounts, Irelied on cognitive appraisal theories of emotion(for a review, see Ellsworth and Scherer, 2003).Numerous empirical studies have established thevalidity of this theory (e.g., Frijda, Kuipers, and terSchure, 1989; Lerner and Keltner, 2001; Roseman,1991; Smith and Ellsworth, 1985; Tiedens and Lin-ton, 2001). Cognitive appraisals of emotions arerelevant to analyzing middle managers’ reportson their experiences with strategy implementa-tion. During private interviews, these managerswere essentially making sense of their reactions toorganizational events deemed important to them.The emotions and events they reported were oftenof an enduring nature and could be reproducedmentally in a relatively reliable manner (e.g.,feeling anger toward newcomer executives whenemployees sensed that their group competencieswere repeatedly not valued by the newcomers), asopposed to less consequential, episodic emotionalexperiences that could be forgotten quickly (e.g.,being upset by a computer malfunction that wasfixed a few days later) (see Robinson and Clore,2002).

Studies of emotion-appraisal associations showthat experiences of specific emotions (e.g., fear,anger) are reliably associated with particular sets ofappraisals (e.g., Smith, 1989; Smith and Ellsworth,1985). These appraisals involve a small num-ber of dimensions such as pleasantness (foster-ing one’s well-being), agency, control, certainty,motive (goal) consistency, and adherence to moralnorms (Ellsworth and Scherer, 2003). For exam-ple, when people feel angry (or remember feelinganger), they tend to report thinking that the sit-uation is unpleasant, that other people intend tocause harm, and that they are certain about whatis happening. This appraisal pattern can be con-trasted with that of fear, for which people reportthat the situation is unpleasant, the causal eventor outcome is uncertain, and they have little con-trol over the situation. Sadness is experiencedwhen people interpret the situation as reducingtheir well-being, the unpleasant outcome is cer-tain but they have little control over it, and thereis no intention by other people to cause harm.Pride is experienced when people appraise the out-come as certain and beneficial to their well-beingand believe the pleasant outcome is caused by

themselves or other people with whom they iden-tify strongly (Roseman, 1991; Tiedens and Linton,2001).

Table 1 shows how group-focus emotions andsocial identities were inferred. The more ofteninvoked identities involve tenure (i.e., length oftime working for the company) and language(mother tongue of English or French). Linguis-tic social identity and the emotion of fear, forexample, were inferred in the following quote:‘The (senior executives) tend to be Anglo-Saxonmen. . .I think French Canadians feel threatened.’In this particular context, the French CanadianHR middle manager described this fear—as anunpleasant feeling state, with low personal con-trol over an uncertain future negative outcome—asa member of the French Canadian group (andon behalf of other middle managers) (group-focus emotion), although he should have had lit-tle objective reason to feel threatened personallybecause newcomer Anglophone senior executiveshad just appointed him to this influential posi-tion. Moreover, virtually all middle managers Iinterviewed still held their powerful positions, butnewcomer top executives left after three years.It was the situational saliency of social identitiesthat elicited their group-focus emotions. Table 2illustrates how middle managers’ inferred nega-tive emotions were supported by company doc-uments such as employee surveys and focus groupreports.

Finally, analyzing interviews with executivesallowed me to infer that although they did talkabout the importance of social-emotional mattersin a conceptual way, they did very little to dealwith group-focus emotions and social identities.What caused this gap between knowing and doingwas a dominant focus on task achievement undertime pressure, although one senior executive didfocus on both social identities and tasks. Table 3illustrates this analysis.

I now describe the findings in a narrative mode.Identifying various emotional experiences basedon narratives that connect expressions of how peo-ple responded affectively to the actions of othersand recording their interpretations and resultingbehavior in a temporal order that suggests a causalrelationship, is commonly used for studying emo-tions in natural settings (Huy, 2002; Locke, 1996;Maitlis and Ozcelik, 2004; Martin, Knopoff, andBeckman, 1998).


1394 Q. N. Huy

Table 1. Illustrations of inferred group-focus emotions associated with tenure and language(FC: French Canadian; V: Veteran)

Quotes Appraisals Inferred emotion Social identity Source

The CEO doesn’t speak French. Sothe Quebec employees don’t feelthat they can have him in theirheart; they can’t relate to him.(Another top team member) doesn’tspeak French. He is responsible for15,000 employees, of which 40percent are in Quebec. . .So theFrench business is now surfacingand becoming quite an issue.

Unpleasant state.Certainty aboutnegativeoutcome causedby others. Lowattribution ofothers’ intentionto harm.

Unhappiness orsadness(negativeemotion)

Language Humanresources(FC)

The fact that they wouldn’t speak verygood French, the French people inQuebec are very tolerant of that.What they don’t expect, and theyfind unacceptable, is that they don’ttry to speak French.

Unpleasant state.Certainty aboutoutcome that isbelow one’sexpectations.

Disappointment(negativeemotion)

Language Humanresources(FC)

If you have more than five years inCanco, you have to prove yourselfanew again. People who wererecruited from outside benefitedfrom much more generouscompensations and they are betterregarded and treated. . .They havea more interesting job. . .Theycommand more attention.

Unpleasant state.Certainty aboutnegativeoutcome.Undeservednegativeoutcome (unfairtreatment).

Anger (negativeemotion)

Tenure Seniorexecutive(V)

French Canadians feel respected in thisinitiative. Our knowledge ofcustomers in the Quebec market wasvalued by the newcomer executive.

Pleasant state.Certainty aboutpositiveoutcome.

Happiness(positiveemotion)

Tenure Salesmanager(FC)

Many of the managers in marketing arenew. Newcomers are more crediblewith senior executives and muchbetter paid than veterans. There’sresentment on this side.

Unpleasant state.Certainty aboutnegativeoutcome.Undeservednegativeoutcome.

Anger (negativeemotion)

Tenure Sales andmarketing(V)

Veterans are so scared to appear tooarchaic or too traditional that we nolonger say what we think. When youhear newcomers say something, youfeel obliged to support them even ifyou think it’s nonsense, because youknow senior management wants newblood and new ideas.

Unpleasant state.Uncertaintyabout negativeoutcome. Lowcontrol oversituation.

Fear (negativeemotion)

Tenure Sales andmarketing(V)

MIDDLE MANAGERS’ GROUP-FOCUSEMOTIONS, ORGANIZATION-RELATED SOCIAL IDENTITIES, ANDSTRATEGY IMPLEMENTATIONOUTCOMES

I will first describe how CEO Foster’s threestructural changes aroused negative group-focus

emotions and caused the saliency of specific socialidentities in middle managers’ minds.

These structural changes involve: (1) changingfrom a function-focus bureaucracy to market-focus profit centers; (2) changing to function-focusdivisions operating as profit centers; and then(3) changing back to a centralized bureaucracy. Allthree changes produced unsatisfactory performancein terms of customer service and profits, leading to



Table 2. Illustration of triangulation of inference of negative emotions with confidential company documents

Company documents Nature of group-focus emotionsNegative emotions are elicited when

people perceive events thatpotentially harm their individual orcollective well-being

Focus groups: many employees believed that Canco was led byoutsiders who were mainly interested in financial results rather thancustomer service.

Tenure (veterans versus newcomers)

Reports from some employee groups: ‘employees feel less valued,respected, and secure because of newcomer executives’ actions.’

Tenure (veterans versus newcomers)

According to 40 representatives of employees, some believed that topmanagers were uncomfortable with the French Canadian culture andso they could not related well with Quebec employees.

Language (English Canadians versusFrench Canadians)

Postmortem report: employees believed that newcomer executives didnot know where the company was headed and failed to react quicklyenough, thus causing underperformance in change effort and muchneedless suffering to veteran employees.

Tenure (newcomer versus veteran)

The company’s random survey of about 2,400 employees conducted inthe third and last year of corporate transformation produced thefollowing results: just more than half (51%) expressed confidence inthe newcomer executive team; only about 40 percent believed thatnewcomer senior management dealt openly and honestly with them;scarcely half (55%) felt valued as employees; only 37 percentbelieved that the company had a sincere interest in their well-being.

Tenure (newcomer versus veteran)

the departure of the top team. I explain why Cancoexecutives were often not able to deal with middlemanagers’ group-focus emotions and social iden-tities as they performed these structural changes. Ithen describe how and why group-focus emotionshad a negative spillover effect on other top-downinitiatives, such as the culture change. Finally, Idescribe one of the rare top-down strategic ini-tiatives that succeeded, in part because the seniorexecutive addressed middle managers’ group-focusemotions and social identities.

First structural change: eliciting negativegroup-focus emotions and social identities

Newly appointed CEO Foster (an English-speakingCanadian) wanted to implement his new corporatestrategy rapidly. He and the top team he appointedbelieved they had to make fast and wide changesto the company’s strategies, structures, culture, andsystems if they were going to achieve alignmentamong key organizational elements to handle thenew environment (Miller and Friesen, 1984). Thefirst of the five top-down initiatives was a changein corporate structure that elicited negative emo-tions among many middle managers.

The first major change enacted by Foster wasreplacing Canco’s divisional bureaucracy with four

strategic business units (SBUs) operating as profitcenters. Foster wanted to reduce managerial com-plexity and devolve more accountability onto eachSBU so that it could become more nimble in fight-ing competition. Foster wanted to improve strate-gic focus with smaller business units that had cleargoals—goals that were profit oriented.

In appointing members of the new senior exec-utive team, Foster believed that many incumbentexecutives were unfit to deal with the new compet-itive context, using the following metaphor in thefirst of three private interviews to describe Canco’schallenging transition from a quasi monopoly to ahighly competitive environment:

‘In the past, we were tied up at the harbor inthis beautiful big luxury liner in Miami. Peo-ple would bring us food; all we had to dowas (shine) the deck and (turn) the motors onand (shut) them back off again. Now, we haveto go to sea. The problem is, not a lot of(the management team) has been to sea before.There are going to be storms and typhoons andenemy gunships. This is a place for survival.I’m going to need leadership that knows howto win the war. . .I’d say that this company was


1396 Q. N. Huy

managerially bankrupt to go where it had to go.’(emphasis mine)

Four of five top executives were newcomers.Combined with the 22 next-level senior executivepositions, about 52 percent (14 of 27) of veteranexecutives were retained in the senior team.6 Fos-ter’s selection criteria emphasized, at least in hisverbal statements, strong task expertise and teamleadership skills. As Foster explained, ‘My phi-losophy about an executive team is that to makesure that each one is best at what they do function-ally—marketing, finance, engineering—they haveto have strong leadership skills, and they have tobe willing to work as a team.’

Tenure-related social identity

As the new corporate structure was being put inplace, many middle managers felt negative emo-tions, such as anger and fear, that were causedby perceived devaluation of their tenure-relatedsocial identity. This was exacerbated when sev-eral newcomer executives belittled veterans (asa general group), publicly calling them unfit forthe new competitive world. Thus, newcomer exec-utives made a valued organization-related socialidentity—tenure—salient to a large number ofveteran managers. The majority of Canco middlemanagers started at entry-level positions and grad-ually, through many years of performance, workedtheir way to higher-level positions; they identi-fied themselves proudly with the rich and diverseexperience they had developed by working for onecompany.

Veteran middle managers interpreted the new-comers’ statements as devaluing their generalgroup organizational competence and became agi-tated, even those individually appointed to power-ful positions by top executives. For instance, oneveteran middle manager in charge of an impor-tant budget allocation in Canco reported in private:

6 I define a veteran as an organization member who had workedfor Canco for at least five years at the launch of the strategicchange. This comes from my understanding of Canco’s culture,which valued long tenure and lifetime employment. This five-year time frame has also been mentioned in interviews withmiddle managers (see Table 3). The majority of veteran respon-dents had worked for Canco for seven to 25 years, and manywere in middle management positions. Newcomers are thosewho joined Canco three years or less before the launch of thestrategic change, including CEO Foster.

‘People who have good knowledge of the organiza-tion become legacy persons. We have two groupsof people now. As a result, the legacy people donot dare to express their opinions clearly for fearthey may be perceived as negative by the newgeneration.’ Another middle manager in an influ-ential change management position confided: ‘Thenew executive was speaking from both sides of hismouth. He went to see one group and said that theother was incompetent. Then he went to the othergroup and told them that the first group was bad.He didn’t know that both of us talked to each other,so the resistance was very strong. You can’t tellpeople they have been imbeciles for years.’ Otherveteran managers felt angry that newcomers haddisparaged their group identity—indiscriminately,frequently, and for too long. People become angrywhen they perceive intentional offense (Stein, Tra-basso, and Liwag, 1993). A middle manager in HRreported what he had observed in meetings: ‘Wewere told all the time (by newcomers), ‘This is theold Canco, this is the new Canco. Everything thatbelongs to the past is considered rotten.’ We. . .hatebeing told continually that we are bad.’

Experimental research has shown that whenmembership in a group experiencing a conflictwith another group is made salient, people dis-play greater identification with the ingroup thanwith the outgroup (Mackie et al., 2000). Thus,although the veteran HR middle manager was notpersonally devalued in his position by newcomerexecutives, he still used the word ‘we’ to describethe shared fate of veterans. The divisive labeling ofnewcomers and veterans fueled intergroup hostility(Bornstein, Budescue, and Zamir, 1997), especiallywhen people perceived ingroup favoritism and out-group discrimination (Miller and Brewer, 1986).

Newcomers’ dismissive treatment of middlemanagers’ tenure thus triggered negative group-focus emotion, as illustrated by quotes shown inTable 1. Many veterans withheld critical piecesof information (such as declining customer ser-vice and rising interunit conflicts) that mighthave allowed newcomer executives to adjust theiractions midcourse. One veteran director of cus-tomer service shared his private views of the newdynamics at formal meetings: ‘Each time we heardan executive or an external consultant suggestsomething, we said we were not going to say any-thing. We don’t want to appear to be dinosaursone more time.’ Middle managers’ negative emo-tions spread through their private storytelling. Fear



may motivate avoidance and anger may motivateconfrontation (Folkman and Lazarus, 1988), butcombined they led to widespread covert resis-tance among Canco’s middle managers. Agita-tion may also cause people to turn their attentionto its source, deflecting that attention from othertasks (Izard and Ackerman, 2000). At Canco, thisdampened collective action to implement the newcorporate structure, as indicated by foot-draggingbehavior in various business units. One middlemanager described it this way: ‘We are now livingin a confederation state. Everyone was in agree-ment in a meeting. . .They agreed they would doit, and as they came out, they would mutter, ‘WhenI want to. And I did not say when, either.”

Language-related social identity

Beyond the social identity of organizational tenurethat vexed many veteran middle managers, nega-tive group-focus emotions among French Canadianmanagers were further increased when newcomerexecutives making change to the SBU structureinadvertently violated another organization-relatedsocial identity: representation of employees ofFrench Canadian descent in senior executive posi-tions. The latter decreased from 55 percent (10 of18 positions before Foster) to 26 percent (sevenof 27 positions). Moreover, these middle man-agers perceived executives’ disrespect for theirsocial identity when executives did not make aneffort to speak with them in French. As one HRmiddle manager reported, ‘Newcomer executivesare Anglophones and do not speak French, ourlanguage. All the communication in Canco is inEnglish, and this is not a political issue, but a mat-ter of respect.’

The language-related social identity needs to beunderstood in context. Over 50 years, as Cancogrew and served markets that used both Englishand French, it created separate divisions to servethem. Company policy was to integrate its ‘minor-ity’ group throughout the firm by including French-speaking employees in middle and seniormanagement positions. This gave them a sense ofpride and a feeling of being treated equally. Bothlanguages were used in official company com-munications. By the mid-1990s, French-speakingemployees accounted for about 40 percent of thetotal workforce. As one ex-CEO told me:

‘The French/English symbols were critical forFrench-speaking customers and that’s for polit-ical and marketing reasons. You have to be closeto the customers you serve. It was a great ral-lying cry for the employees of French Cana-dian ancestry. Employees recognize themselveswithin the company in which they work.’

Only one person on the new top team of five mem-bers, Francois Tremblay, was both a veteran andFrench Canadian. Foster’s declared selection cri-terion was functional-technical skills, but othersinterpreted his choices differently. French Cana-dian middle managers and union officials per-ceived Tremblay as a token with little clout. Theirdissatisfaction was never made public, as invok-ing minority representation had previously beenconsidered formally irrelevant to appointment tomanagerial positions. The company’s ongoing offi-cial policy held that such appointment should bebased on task competence and fit.

In public, French Canadian and veteran man-agers acted calmly. In private interviews, however,many managers expressed negative group-focusemotions. For example, one veteran French Cana-dian senior middle manager in human resources(who had been appointed to this influential positionby newcomer senior executives) reported: ‘Rightnow in this organization, it’s very dangerous andvery unfashionable to be a French Canadian and,especially, a French Canadian who complains.’This well-networked HR middle manager reportedthat she had noticed the same apprehension amongcolleagues who formed part of her informal privategroup. (Table 1 shows some illustrative quotes.)Predictably, opposition behaviors arose.

Covert conflicts between veteran and newcomermanagers slowed implementation. One middlemanager reported foot-dragging by his colleagues:

‘One (newcomer) senior executive wanted a cer-tain change to be implemented in the company,so all the managers around the table said, ‘Yes,no problem.’ Two months later, he asked them,‘How come it’s not done yet?’ They just told him,‘Well, it’s not quite in place yet.’

Another veteran French Canadian middle man-ager explained, ‘When people don’t agree with adirective, they just ignore it. It’s faster to ignorethan to argue with a decision. Or they would justsay, ‘This is not quite what we agreed on.” Another


1398 Q. N. Huy

French Canadian middle manager observed, ‘New-comers have difficulty operating because theydon’t know how things work around here.’ Inter-nal conflicts contributed to declines in customerservice and market share, leading CEO Foster toconclude, only 15 months after its introduction,that the first structural change had not produced theexpected performance improvement and that, onthe contrary, he saw deterioration. To fix this prob-lem, he decreed further structural change. Whenasked privately the reasons for this change in cor-porate structure in the second year of CT, Fosterattributed the underperformance mainly to task-related causes:

‘What I found was, we were beginning to spend atremendous amount of our time discussing inter-nally with each other instead of focusing on theexternal customer. . .Secondly, with a companyas integrated as ours is, it became much morelogical to operate this on an integrated basis asone operation, but with very distinct functionalresponsibilities, as opposed to distinct, separatecenters. So we reintegrated the company aroundwhat we call the customer-facing organization,and then we reassigned some responsibilities.’

Second and third structural changes: blind tomiddle managers’ group-focus negativeemotions and continuing to unintentionallyviolate their social identities

The second structural change involved the creationof a hybrid division bureaucracy that included aCEO and executive vice presidents, each of whomacted as president of a profit center that coveredoperations, sales and marketing, corporate admin-istration, and innovation (this division along tra-ditional functions can be contrasted with divisionalong different customer markets in the first struc-tural change). The second structure lasted for abouta year. The percentage of veterans on the seniormanagement team dropped from 52 percent to 43percent (12 of 28 positions), while the percentageof French Canadians on the senior managementteam dropped from 26 percent to 21 percent (sixof 28 positions).

The middle managers’ emotional responses—linked in part to group-social identities—remained.In terms of organizational effectiveness, underper-formance problems persisted, and this led Fosterto order a third change in corporate structure, in

which the percentage of veterans in senior man-agement dropped from 43 percent to 37 percent(nine of 24 positions); the percentage of FrenchCanadians remained unchanged at 21 percent (fiveof 24 positions).

This change consisted of creating a bureau-cracy. One top team member was fired and theonly veteran French Canadian top team membersaw his responsibilities reduced to public relations.The third corporate structure lasted for only sevenmonths, at the end of which Foster and the topteam resigned. In sum, under Foster’s leadershipof Canco, the three corporate structure changessaw veteran representation on the senior manage-ment team drop from 94 percent to 37 percent andFrench Canadian representation drop from 55 per-cent to 21 percent.

These successive changes in the percentage rep-resentation of veterans and French Canadians insenior management positions can be linked to anintriguing phenomenon called ‘people accounting.’Recent experimental research has shown that deci-sion makers can make resource allocation choicesin highly competitive situations based on theirconsideration of the representation of social cat-egories, as opposed to individual factors such aspersonal competence and merit (see Garcia andYbarra, 2007). Perceived numerical imbalances onsocial identity factors that are meaningful to peo-ple can prompt their fairness concerns to shift fromthe interpersonal to the intergroup level.

Thus, Canco middle managers were concernedabout either their tenure- or their language-relatedgroup’s percentage imbalances in the senior exec-utive positions. Regarding tenure, one HR middlemanager noted, ‘People here felt an incredible frus-tration, not because we hired newcomers—it isnecessary to hire newcomers—but people haveproblems with the sheer number of them.’ Asfor language identity, the percentages of Frenchand English senior executives in Canco’s corpo-rate structure reflected, in part, the larger English-French power balance that characterizes Canada’shistory. French Canadian middle managers con-sidered these positions as meaningful symbols ofFrench Canadians’ representative power and dis-tributive justice, even if the majority of themwould never be appointed to the top positionsthemselves.

The negative group-focus emotions and deval-ued middle managers’ social identities reduced theimplementation success of not just the initiatives



that triggered them, but also other initiatives, evenif the latter seemed intrinsically desirable to mid-dle managers. I call this the ‘emotional spillover’effect.

Emotional spillover: dismissing a culture of‘warm intensity’

CEO Foster was not totally insensitive toemotional-cultural issues. In parallel with strate-gic initiatives to change the corporate structure asdescribed earlier, Foster wished to change the qual-ity of working relationships inside Canco—thecultural ‘software.’ He envisioned a climate of‘warm intensity,’ as he explained in his first privateinterview:

‘Intensity means you are very focused, there’s anurgency to everything you do. . .Intensity is veryconstructive and positive. . .Warmth is creatingan environment where we are deeply commit-ted to each other, to our company, and to ourcustomers. And we go through the tough times,the down times, together. . .I wanted to be veryhonest with everybody and tell them that thesechallenges and the change we are about to gothrough will be dramatic; it will impact everyone of you. But it can be okay. We can getthrough this. . .I do not believe in fear. . .I don’teven believe in dictating.’

Although the declarative content of the proposedchange seemed desirable, it evoked little excite-ment at lower levels of Canco. Middle managersremained indifferent, cynical, or apathetic. Thesefeelings could have been caused in part by theirperception that the top team had initiated very fewconcrete and credible actions to realize the desiredculture change. Very few interviewees mentionedthis change without being specifically prompted.When asked, they responded with emotions suchas disappointment and contempt. The top team pro-vided anything but role models of warm intensity,they said. Several respondents used the term ‘hotintensity’ to describe the temperature of the covertopposition in various work units.7

7 Middle managers’ sentiments seemed to be largely shared bylower-level employees. A random survey of about 2,400 employ-ees (which included middle managers’ responses), conductedduring the third and last year of strategic change, produced thefollowing results: about 60 percent believed that senior man-agement did not deal openly and honestly with them, about

Negative emotions accompanied middle man-agers’ questioning of the authenticity of the ini-tiative and prompted dismissive behavior. As mid-dle managers experienced negative group-focusemotions toward the top team in part because ofdevaluation of their social identities, their emo-tional resentment spilled over to dismiss a strate-gic initiative (warm intensity). This initiative wasimplicitly linked to these identities, as middle man-agers understood that ‘warm, deep commitmentto each other’ meant that they, and their tenureand language social identities, should be valuedand respected). Negative group-focus emotions inresponse to this initiative undermined its imple-mentation success right from the beginning. Manymiddle managers treated the proposed change inthe quality of relationships as just another emptypiece of corporate rhetoric. The attempt to changethe firm’s culture petered out, leaving the CEO toponder, ‘One of the great challenges is how to keepthe troops motivated.’

This case also shows how strategy implementa-tion can derail without the top team being awareof it. CEO Foster was undoubtedly sincere in hisdrive to create a warm relationship culture, buthis implementation actions were perceived by mid-dle managers as displaying disrespect for theiremotion-laden social identities. The perceived gapbetween the CEO’s rhetoric (his intention) and hisactions (implementation process) generated muchnegative group-focus emotion and mistrust thatultimately led to the failure of this cultural ini-tiative, displaying patterns in middle managers’group-focus emotions and covert opposition sim-ilar to those that occurred with the failed imple-mentation of the three corporate structures.

These cases show the critical contribution ofmiddle managers to strategy implementation.Implementing a new corporate strategy requiresmiddle managers’ enthusiastic endorsement, whichenergizes them to go beyond the call of dutyto resolve unforeseen problems that surface dur-ing implementation (Simonin, 1997). Because they

45 percent did not feel valued as employees, and only 37 per-cent believed that the company had a sincere interest in theirwell-being. Furthermore, about half (49%) expressed low to noconfidence in the top team. These surveys were commissionedby Canco’s executives and were conducted by telephone. Asemployees risked being identified through their telephone andemployee numbers, some caution should be exercised in inter-preting these results. Given the climate of fear, downsizing, andmistrust, I speculate that the survey results may be biased infavor of the executive team.


1400 Q. N. Huy

have control over important organizationalresources, including the many people reporting tothem, middle managers’ indifference, apathy, pas-sive resistance, or even mechanical compliancemay be sufficient to undermine the requisite col-lective energy for improvisation, creativity, abovenormal efforts, and persistence—all of which con-tribute to the success of ambitious changes inan organization (Huy, 2002). Aggregate businessunderperformance can be caused by a small num-ber of work units (led by reluctant middle man-agers) affecting interdependent units (Thompson,1967).

Top executives’ task-focused behaviors intop-down strategy implementation

Analysis of the senior executives’ interviewsthroughout all three years of corporate transfor-mation suggests a dominant action focus on taskmatters that crowded out their concerns aboutsocial-emotional issues. This does not imply thatsenior executives did not consider social-emotionalmatters as they planned strategy implementation.Senior executives did talk about the importanceof generating pride among employees, creatingwarm intensity, and building cohesion and shar-ing of knowledge among the senior executives.Unfortunately, they did not find adequate timeto act on social-emotional matters, as they wereoverwhelmed by numerous challenging tasks andfelt intense pressure to meet deadlines and finan-cial goals. In other words, there was a large gapbetween ‘knowing’ the theoretical importance ofsocial-emotional matters and actually devoting sig-nificant resources to knowing about and addressingthem (‘doing’). When asked in the second yearof CT what his key priorities were, CEO Fos-ter summarized his focus on organizational struc-ture, corporate strategy, and meeting importantconstituencies:

‘There are (several) major initiatives we haveto do as a company to get better. (The) first onewill be a very, very focused, realigned structure,and I’m working on realigning the structure ofthe company. . .Second, also tying that (struc-ture) together with where we believe the wholeenvironment is going. . .Third, we are going torevector our strategy. . .We are going to restruc-ture so that we really organize our company tomeet customers’ requirements first. I’m working

very hard on this right now. . .Fourth, I am goingto share with the government where this industryis all going and develop a vision of the indus-try in Canada for them and how they can meettheir policies and give us a chance to succeed. . .

Fifth, I’m certainly going to go to all (middlemanagers) over the next month. Sixth, I’m tak-ing the board of directors away for a day and ahalf, and I’m sharing with them where this is allgoing.’

As this quote suggests, Foster intended toannounce the new structure and strategy to mid-dle managers so they could be informed and getaligned to execute the new strategy and structure.His intent was not to explore middle managers’feelings. Foster seemed to expect that the seniorand middle managers he appointed to leadershippositions would execute his strategy. He did notsuspect that these managers had negative feelingsabout his new structure:

‘I. . .technically put teams in place (with individ-uals) from inside or outside of Canco that arecapable of doing (their jobs), and I expect thatthey all will deliver. . .My role now becomes oneof managing from altitude the work. . .explainingto government—that’s as big as any other singlechallenge. . .talking to the investors, and tryingto keep myself up to date with where the industryis going from a technology market and organi-zational point of view. That keeps me fairly wellemployed 150 percent of my time. One of thethings that I get disappointed in is that I don’tget back to talk with the employees.’

Additionally, as described above, Foster’s dom-inant selection criterion for new executives joiningCanco was task competence. At no time was theissue of English-French Canadian social identitiessalient in his selection criteria for appointing newsenior executives:

‘They had to be very, very good functionally attheir job. Number two, they had to have strongleadership skills. They had to be able to reallystand up and demonstrate leadership when thegoing gets tough. Number three, they have tobe willing to work as a team, where the goalsand objectives of the team come ahead of theindividual. That is not trivial.’



Table 3 contains further supporting evidencebased on private interviews with other seniorexecutives. Senior executives were also so over-whelmed by their task challenges that they didnot have time to interact more often among them-selves to build a more cohesive team, even ifthey recognized the importance of doing this ina theoretical sense. Under resource constraints,social-emotional matters always had lower priori-ties than tasks. Consistent with this task-orientedschema, the CEO and his top team sought to solveorganizational effectiveness problems by recruit-ing even more outsiders to fill influential positions,thus exacerbating the divide between veterans andnewcomers and amplifying dysfunctional group-focus emotional cycles caused by violation ofmiddle managers’ social identities (see quotes inTable 3 that show how newcomer senior execu-tives emphasized new task competencies that onlynewcomer managers could bring to Canco). Rais-ing soft issues in a high-pressure task-focusedclimate seemed unwelcome. As one veteran HRmanager who dealt frequently with senior execu-tives noted:

‘You have a very short window of opportunityto talk about anything that’s called ‘organi-zational culture’ in this company. The higherup you go, the less they want to hear aboutit. . .Their eyes glaze over very quickly when youstart talking about nontangible things. . .A lotof people don’t want to hear that you have tolook at systemic causes behind behavior. Theythink it’s too theoretical. . .Their natural wayof behaving and making decisions and decid-ing what’s needed would not predispose themto realizing the importance of the human sideof management. . .Many of the newcomers atthe top of the company do not understand orvalue what employee commitment is all aboutand how it has a direct link to how profitableyou’ll be and, over time, how successful you’llbe.’

Not all of the top-down strategic initiativesunderperformed. The next case shows how a seniorexecutive can address group-focus emotions andsocial identities constructively.

Addressing middle managers’ group-focusemotions and social identities

The industry specialization in sales (ISS) initia-tive was launched by another Anglophone new-comer executive. Ted Smith was in charge of about3,000 employees in sales and marketing (SM).He wanted to create industry segments such asfinancial services and airlines to increase customerfocus and gain deeper knowledge of various indus-tries so as to serve them better. Industry specialistswould replace generalists, who had proliferatedunder Canco’s universal customer service mind-set. In the past, Canco’s SM group had beendivided along linguistic lines to serve customers ineither English or French, but Smith believed thisstructure created costly duplication and too manydivergent approaches. He believed combining thelanguages and dividing the SM group by industrysegment would improve profitability by spreadingthe development and marketing costs of new prod-ucts across a larger mass.

As could have been expected, the proposedchange initially elicited negative emotions amongFrench Canadian middle managers because thenew industry structure would no longer recog-nize the distinctiveness of the French Canadiansocial identity, but this group-focus emotion wasanticipated and addressed quickly. Fortunately, aconsulting firm brought in by Smith to imple-ment his plan had worked with Canco veteransin the past and was well aware of the company’sidentity sensitivities. The consultants proposed animplementation process that addressed the twoorganization-related social identities: tenure andlanguage.

In the planning phase, consultants assisted man-agers in their deliberation and provided informa-tion about information technology markets. OneFrench Canadian middle manager reported: ‘Con-sultants also played the role of confessors; youcould use them to relay your messages upward anddownward.’ Open exchange with veterans allowednewcomers to become more aware of Canco’semotion-laden social identities, and this allowedthem to adjust and frame their implementationapproach. ISS was framed as eliminating separateEnglish- and French-speaking divisions in orderto improve customer service, which was valuedby veterans. As a result, the proposed change waswell received by many middle managers.


1402 Q. N. Huy

Table 3. Illustrations of analysis of interviews with senior executives

Top executives’ quotes in private interviews First-order inference Second-order inference

Top executive (interview in the middle of final year of CT)If we’d been together five years ago, we would have done

something different two years ago than we did. Time is afactor. Time is a critical factor and we didn’t have a lot.$900 million, $700 million, going to $500 million net incomecatches your attention. . .So the company was completelyfocused on making its financial targets. So many things got lostin the shuffle of the necessity to make the financial targets.

Time constraint andfocus on financialresults

Time pressure withtask focus

Canco people’s. . .experiences were all under one set of rules withone set of orientations and, frankly, a very narrow view of theworld. . .It was rather experience that we went to buy in themarketplace. . .Knowledge of the consumer market, togetherwith leading-edge marketing and sales capability will, in fact,be what will succeed in the future. . .We needed people withexternal skill sets who understood that, who were recruited.

Selection of newmanagers based ontask requirement

Task focus

We learned a lot about our limitations of our informationtechnology capabilities through this period. It’s a problem weshould have known before. . .Our dependence on IT is sohuge. . .But we only knew about six months ago that thecapacity of IT is not going to be enough.

Focus on informationtechnologyproblems

Task focus

Senior executive (private interview mid-second year of CT)I want to manage with a strong, strong orientation toward product

management. The company needed a management system thatwould drive cost out of the business. I had no idea about howmuch anarchy there was, how much lack of discipline therewas, how difficult it would be.

Focus on financialresults

Task focus

Among the (newcomer) senior vice presidents, (we would haveliked to have had) a chance to know each other a little bitbetter and develop a little bit more mutual respect for eachother. But there’s no other medium, other than officer meetingsthat occur biweekly. . .Because we’re all so new with eachother, (there is) a subtle culture of conflict avoidance. (Thisavoidance exists) among the top team and the other 20 seniorVPs and any of the 20 in and among themselves. Also, therehadn’t been a sufficient number of instances where we couldtangibly demonstrate our support for each other. . .I think thatthe officer team needs to learn how to have fun together, andthey haven’t done any of that.

Focus on task;neglect importanceof developingcohesive seniormanagement team

Time pressure andtask focus

Senior executive in charge of successful ISS project (interviewfirst quarter after end of CT)

Yes, it was reasonably easy because we worked closely togetherand made sure that the interests of both parties were taken careof, and we put leaders in place that equally represented(English) Ontario and (French) Quebec. The concern is moreQuebec, and it was handled fairly well. We also consolidatedbuying power and started to centralize buying decisions, and wecentralized how we interface with our business partners.

Social identity(language) and taskfocus

Address socialidentity and taskfocus

What happened is some of the customers who do a lot of businesswith us were finding we weren’t doing a lot of business backwith them. So we had to shift that back a little bit, and wemade changes where we put in a president of Canco Ontarioand a president of Canco Quebec, and that was to focus moreon how we interface with our customers in each of the twoprovinces. It didn’t change the industry focus, which was leftintact. So that change was deemed to be a good one.

Social identity(language) and taskfocus

Address socialidentity and taskfocus



Smith set up 12 implementation teams andappointed team leaders who represented both lin-guistic groups. He had equal English- and French-speaking representation at the level of team lead-ers. As one manager described it:

‘We mix, right at the design stage, salespeoplewho have been working in silos in different divi-sions, and [we do] this at all levels. It was noteasy for us initially, because we have been work-ing independently for the last 15 years. At thestart, you felt that the other side did not reallyrespect you because of your linguistic identity.Teams often had a dual prime, each representingone linguistic group.’

This dual leadership created positive emotionsamong French Canadian managers and increasedtheir receptivity to this initiative. As one veteranFrench Canadian middle manager said:

‘Having separate sales structures in differentgeographical areas for large and medium-sizebusiness customers is just too costly, too slow,and we cannot respond fast enough at thenational level. Thus, we have to integrate vari-ous divisions. Newcomers and consultants gavethe opportunity to everyone to express them-selves and to be listened to. Everyone was look-ing for a means to accomplish this change in avery short time.’

Smith also met with many veterans to learn fromthem and to demonstrate that their opinions andsuggestions were factored into the implementationof change (e.g., dual team leaders). Positive emo-tions were apparent in the words of two veteranmiddle managers who were invited to lead thechange team and who belonged to the French- andEnglish-speaking groups respectively:

‘I think this change is good. We cannot letpeople stay mired in their old habits; we needto shake things up. . .We are putting a solidstructure in place, with people specializedin various industry segments, with increasedteam focus on particular needs. This allowsme to better understand the market.’ (French)

‘With ISS, and as opposed to other (major)changes, (the newcomer executive and con-sultants) did listen and opinions were

respected. . .Around 200 middle managersworked feverishly in a change mode for sev-eral months. I think this is a very good ini-tiative. We will be able to give a hard time toour competitors.’ (English)

Shared positive emotions allowed openexchanges of interim results and rapid adaptationto unforeseen problems. For instance, about twoyears into implementation, sales middle managersfound out that too much consolidation had dam-aged sales. Several large French-speaking clientsexpressed discontent at the reduction in contactsthey had with executives and at the reduced atten-tion to their particular needs. These customers’negative emotions were quickly relayed to Smith,who, acting on advice from middle managers, cre-ated the ‘front end’ positions of ‘president, Englisharea’ and ‘president, French area’ for customerrelations. A matrix structure was put in placewhereby a layer of English- and French-speakingcustomer-representative groups overlaid the verti-cal structures. Hence, the voices of the English-and French-speaking customers could again beheard inside Canco at the executive level (seequotes at bottom of Table 3).

These flexible midcourse adjustments were seenas valuing middle managers’ social identities, andthe initiative was implemented successfully. Therewas also external validation: outsiders noted apromising improvement in the company’s abilityto fight back on the sales and marketing side. Oneprominent bond rating agency noted: ‘Canco hasspent considerable time and resources to becomea more effective marketer. These marketing pro-grams have slowed market share losses and helpedwin back customers.’

This case shows how a newcomer executivecould mobilize veteran middle managers to sup-port a top-down initiative by leveraging their socialidentities. Consultants who had previously workedwith these managers helped this Anglophone exec-utive expand his implementation schema beyondan affect neutral task focus. It follows that topexecutives can extend their abilities to understandand deal with middle managers’ tacit emotion-laden social identities by working with emotion-sensitive advisers and that, given time constraints,this would seem more realistic and efficient thanhaving them improve their personal emotionalintelligence abilities.


1404 Q. N. Huy

It is intriguing that the lessons of this suc-cess seem not to have been shared with otherexecutives. The analysis summarized in Table 3suggests that Canco executives were focused ontheir own numerous task priorities, which mostdescribed as ‘time consuming’ or ‘draining.’ Theyfelt time pressure to meet financial targets in threeyears, and top executives’ individual task focusmight have given them very little time to reflecton their experiences, draw lessons from them,and share them with other executives. As CEOFoster recognized, ‘I didn’t stay focused enoughwith the senior executives so that I (could) con-sistently impart to them how I thought it shouldgo, or get feedback from them. I left them ontheir own. We were not involved in it togetherenough.’

CONTRIBUTIONS TO THEORY,PRACTICE, AND FUTURE RESEARCH

This research started by noting that the top-down strategy implementation literature has gen-erally adopted an affect neutral task, or cognitivefocus. Middle managers’ opposition has often beenattributed to cognitive inertia or the defense ofnarrow self-interests (e.g., individual status or eco-nomic gains). In addition, it seems intuitive thatmiddle managers experience individual emotionsas they appraise a proposed strategy as helpingor harming their self-interest. This three-year real-time field research reveals, however, that somemiddle managers who were appointed to influen-tial positions by senior executives—and so shouldhave perceived low direct threat to their per-sonal privileges—experienced group-focus emo-tions linked to organizational social identitieswhich, in this particular context, involved tenureand language.

This study enriches psychological research instrategic management by showing howorganization-level actions (senior executives’ strat-egy implementation actions) can elicit individual-level, privately felt group-focus emotions that areassociated with middle managers’ social identi-ties (group-level constructs) and organization-leveloutcomes (strategy implementation success). Inaddition, this study contributes to other literaturesdiscussed next.

Organization-related social identity as symbol

This study also enriches the social-psychologicalliteratures it draws from. Although the socialidentity literature has highlighted society-relatedidentities such as profession and gender, this studyadds to it by highlighting the symbolic importanceof organization-related social identities for middlemanagers as key contributors to top-down strategyimplementation outcomes. Many ‘social facts’ canhave intrinsic and symbolic meanings (Zott andHuy, 2007). The intrinsic dimension correspondsto objective or tangible functions that are recog-nized independently of the symbolic dimension.For example, the size of the economic incentives amiddle manager gets is relatively objective. Mean-while, the symbolic dimension refers to evokedmeanings: people make inferences on the basis ofshared interpretations, which can vary significantlyfrom one cultural milieu to another.

In the case of Canco, the long tenure of vet-eran managers was imbued with the symbolicmeanings of organizational experience and com-petence, while French Canadians’ representation insenior management positions and senior executivesspeaking French to French Canadian employees(even if not well) were interpreted by the latter assymbolic acts of respect for their social identity.Although a very small fraction of Francophonemiddle managers would have actually attained thefew senior posts, this symbolic representation wasnevertheless critical to securing the goodwill andsupport of a large number of French Canadianemployees.

The emotion-arousing symbolism that occurredat Canco was not unique. When Jack Welch closedthe small refrigerator division on which GeneralElectric (GE) was originally founded, he discov-ered, to his surprise, that this action aroused moreintense group-focus emotions across the whole GEthan cutting 100,000 jobs. Although top manage-ment found this group-focus emotion irrational,widespread agitation and opposition lengthened theprocess of implementation and made it very costly(Tichy and Sherman, 1994). These authors notedthat Jack Welch learned from bitter experience‘how profoundly emotions can influence humanthought and behavior—and entire organizations’(Tichy and Sherman, 1994: 63) and observedthat ‘most organizations don’t know how to dealwith emotions, so they try to pretend it doesn’texist. By design, corporations seem emotionally



barren’ (Tichy and Sherman, 1994: 64). Again,this group-focus emotion can be explained viaorganization-related social identity.

Language as symbol

Language as a potentially potent emotion-arousingsymbol in strategy implementation can bear impor-tant implications for the success of other kindsof changes, such as joint ventures or mergers andacquisitions (M&A) in multilingual, multiculturalfirms. A recent meta-analysis of 46 studies onM&A suggests that cultural differences can createmajor obstacles to achieving integration benefits(Stahl and Voigt, 2008). My study linking mid-dle managers’ group-focus emotions and linguisticsocial identities could add to a richer and moreprecise understanding of specific factors that havebeen lumped together as cultural differences.

Although the French-English issue is a specialcase stemming from the bilingual nature of manyCanadian firms, one can hypothesize analogouschallenges in some U.S. firms with a high num-ber of Hispanic middle managers, for example.Recent field research has suggested that languageas a symbol of power in corporations is salient inEuropean firms. For example, Vaara and colleagues(2005) described how language skills became anemotion-laden confrontation between managers ofa Swedish bank and a Finnish bank that mergedwith each other. Many managers of the smallerFinnish bank resented and strongly opposed thefact that they were required to speak a foreignlanguage (Swedish) of which they did not havefull mastery, leading them to experience inferi-ority when expressing their perspectives to theirSwedish counterparts, who could communicate intheir mother tongue. Thus, language skill becamean instrument of power and symbolized perceivedinjustice. In the end, English as a foreign languagewas used as a symbol of equality among employeesof the two Scandinavian banks.

Language skill did not seem to be an issueat Canco because many French Canadian middlemanagers spoke English fluently (although the con-verse did not apply to English Canadians). Still,mother tongue can represent the heart and symbolof specific national and ethnic cultures and canarouse strong group-focus emotions on both sidesof the language divide. It can even transcend com-pany levels of analysis, as evidenced by the con-troversy surrounding Spanish as a potential second

official language in certain U.S. states (Schmid,2001).

Group-focus emotion

This research also contributes to the group-focusemotion literature (Smith et al., 2007). It tenta-tively suggests that middle managers’ group-focusemotions can act as manifestations of relevantorganizational social identities that need to beaddressed to enhance the odds of successful imple-mentation. In a large firm, it may be more practicalfor top executives to monitor patterns of group-focus emotions among middle managers (who arein charge of various work groups) than among allemployees. This practice may provide executiveswith early important feedback on whether theirimplementation is progressing smoothly and allowfor midcourse adjustments.

Middle managers, like all individuals, are likelyto have many social identities (Garcia-Prieto, Bel-lard, and Schneider, 2003), and it can be difficultfor top executives leading strategy implementa-tion to predict which ones matter in a particu-lar situation. Because emotions are aroused onlyabout issues people consider relevant and impor-tant to them in a given situation (Lazarus, 1993),the identities that matter most to people in a par-ticular situation will be the ones that elicit emo-tions, positive or negative. These emotion-elicitingidentities should be subjects of open and safe dis-cussion between executives and middle managers,so that they can jointly understand the ones thatreally matter and need to be addressed. Group-focus emotion, thus, represents a situated markerof group identity salience in a particular situation.

Participation and voice

This study provides new boundary conditions thatdeepen our understanding of the social-psychological mechanisms that underlie otheremotion-laden processes—such as participation,voice, and perceived procedural fairness—thatdetermine people’s willingness to support a pro-posed change (See, 2009). One can speculate thatsome nontraditional knowledge content, such asgroup-focus emotions and social identities, maynot surface even in contexts in which participationand procedural fairness prevail, for several reasons.First, some people may feel general negative feel-ings but not be able to articulate them or identify


1406 Q. N. Huy

the specific causes (Seo and Barrett, 2007). Sec-ond, if all organization members in leadership posi-tions are deeply socialized in norms of conduct thatemphasize affect neutral task-focused schemata,they may adhere unquestioningly to these normsand authentically believe that it is not professionalto voice emotional and private matters for opendiscussion at work (Sanchez-Burks, 2005).

CEO Foster’s low priority in dealing with mid-dle managers’ group-focus emotions may be inter-preted as stemming from the influence of theProtestant relational ideology (PRI) work ethic,which holds that effective management of firmsrequires restricted attention to emotional and rela-tional matters (Sanchez-Burks, 2002). This patternof behavior does not need to involve individualemotional intelligence; people influenced by PRIcan competently perceive and deal with the emo-tions of relatives and friends outside the work-place. But many executives continue to believe thataffect neutrality and task focus are the best waysto deal with emotional situations. They avoid dis-cussing emotions at work and understanding theircauses. Consequently, emotions are driven under-ground to incubate and surface later (Maitlis andOzcelik, 2004).

Canco middle managers were most likely social-ized to believe that promotion was based on merit(through annual evaluations of their work perfor-mance and leadership potential), so they mighthave considered it unprofessional and illegitimate,for example, to express openly to Foster that theywere privately bothered by the low representationof French Canadians in senior management. Fosterand other Anglophone newcomer senior execu-tives (as shown in Table 3) did assert in privateinterviews that they used task competence as themain selection criterion. Moreover, middle man-agers might be concerned that top executives coulddoubt their leadership ability and motivation (asthese managers had been appointed to influentialpositions by newcomer top executives) if they wereto display negative emotions in connection withchange or express personal concerns that were notlinked to organizational effectiveness. Argyris andSchoen (1978) and Schein (1992) have noted thatshared tacit, unquestioned assumptions about whatis appropriate for discussion at work will dividework matters into legitimate versus taboo sub-jects. As a result, managers may not express emo-tional matters or put them on the business agenda

even if wide participation, procedural fairness, andempowerment are practiced.

Fostering increased organizational attention tosocial-emotional matters

What could help break this barrier, then, is leadersactively encouraging the expression of emotionsand their causes at work, and this expression ismore likely to occur in a climate of psychologicalsafety (Edmondson, 1999). Seo and Barrett (2007)suggested an alternative approach organizationscould take to feelings: fostering the freedom ofmanagers and employees to experience and expresstheir emotions to maximize the positive outcomesof those feelings, while simultaneously helpingthem minimize the emotions’ potential negativeimpacts. Obviously, any type of freedom has to bebounded: freedom of emotional expression needsto be balanced with respect of other individuals’sensitivities and the company’s interests (Martinet al., 1998).

More specifically, managers and leaders can cre-ate norms of experiencing and expressing a widerange of emotions and their causes by carefullyreexamining taken-for-granted beliefs, languages,and practices that devalue, discourage, or con-strain those feelings (Ashforth and Humphrey,1995). Actively removing cognitive, normative,and behavioral barriers in organizations willrequire much reeducation and unlearning. Man-agers can also increase their emotional self-awareness by acquiring a richer vocabulary ofemotions—by understanding the causes and con-sequences of various emotions such as shame,guilt, anger, pride, and joy so they can recognizethem, regulate them, and express them to others inan articulated way.

More generally, paying attention to patterns ofemotions may allow change leaders to understandthe causes of employees’ reactions to a proposedchange and to address these if necessary. Sanchez-Burks and Huy (2009) argued that change leaderswho have a high ability to accurately perceive pat-terns of emotions in a collective (they call this abil-ity ‘emotional aperture’) may have a higher abilityto manage strategic change. Imagine, for exam-ple, that in response to the company’s announcedplan for a disruptive change, a leader perceives that80 percent of people in her business unit experi-ence fear and the remaining 20 percent experiencehope. Whereas the experience of fear generally



stimulates risk-averse behaviors that likely slowthe implementation of strategic change as peopleexperience lower levels of personal control, experi-encing hope can prompt energetic collective mobi-lization for change. The leader can, for example,encourage this minority group to share the reasonsfor their cheerfulness with their fearful colleaguesand foster an open and honest debate about the per-ceived benefits and costs of the proposed change.In this way, perceiving distinct patterns of group-focus emotions that compose a group’s affectivedistribution can enhance the leader’s change effec-tiveness more than perceiving just a single collec-tive’s mean or modal emotion.

Senior executives could, thus, take certainactions to foster increased organizational atten-tion to social-emotional matters. Examples of theseactions include: foster collective discipline amongthe senior executive team to allocate sufficient for-mal time to discuss soft cultural, social-emotionalmatters because they matter to organizational effec-tiveness, especially when intense stress and pres-sure on task results is felt; assign a well-respectedsenior executive to play this enforcement rolewith other executives; continually develop skillsin emotional awareness among top executives anddiffuse emotion awareness training among mid-dle managers; increase skills among managers toverbalize, display, and distinguish various emo-tions, their likely causes, and consequences (Seoand Barett, 2007); develop increased awarenessand accurate perception of shared emotions thatis distinct from individual, interpersonal emotionalperception (Sanchez-Burks and Huy, 2009); createand diffuse norms of expressing and discussingemotions at work, in a way that is respectfulof and sensitive to the personal needs of oth-ers; be role models and reward other employeeswho perform these behaviors (Martin et al., 1998);create an organizational climate in which peoplefeel safe to discuss sensitive issues (Argyris andSchoen, 1978; Edmondson, 1999); develop, main-tain, and continually refine organizational routinesthat enable managers to perceive, discriminate, andattend to other employees’ emotions when neededto achieve organizational change and renewal. Forexample, managers can organize meetings thatare led by competent and impartial facilitators inwhich employees can be coached to (1) recognizeand express their and others’ emotions; (2) knowwhen and how to express them (e.g., away fromhigh task pressure moments); or (3) deal with these

emotions in a culturally appropriate manner (Huy,1999, 2002).

RESEARCH LIMITATIONS ANDFUTURE RESEARCH

Generalizing findings from this one company studyshould be treated with caution. First, group-focusemotions and social identities (including languageand tenure) represent some of the potential fac-tors that can help or hinder the implementationof strategic change, but they may not be the mostimportant factors for all kinds of change initiativesfor this company or for other organizational con-texts. But the very limitations of this study repre-sent opportunities for future research. Only a care-ful understanding of a number of factors, includingthe organizational context, various kinds of changeinitiatives, and interactions among various groupsof actors, will allow future researchers to establishthe relative importance of group-focus emotionsand various social identities that influence mid-dle managers’ behaviors in regard to implementingstrategic change. Future research could investigate,for instance, which social identities become salientand arouse strong group-focus emotions in var-ious contexts. For example, the social identitiesof physicians and nurses are likely to be salientin a medical context; the identities of adminis-trators and tenured versus nontenured faculty arelikely to be salient in a university context. Scholarscan also investigate the causes and consequencesof language conflicts that arise in mergers andacquisitions that involve companies that previouslyoperated in only one language.

I hope this study has shown that we can gen-erate fresh insights by studying jointly organiza-tion members’ emotions, cognition, and behav-iors that arise during strategic change, rather thanmaintaining the traditional divide between cogni-tion and emotion. In particular, research on strat-egy implementation can benefit from investigatingboth the tangible, rational aspects of organizationaldesign, structures, and processes as well as themore intangible, hidden social-emotional aspectsof managerial behavior. As this study shows, thestructural, social-cultural, and emotional dynam-ics of strategy implementation are interwovenwith one another in organizational settings andneed to be understood holistically. Managerial


1408 Q. N. Huy

belief systems about what is more important (typ-ically the tangible aspects) and managerial limitedattentional resources and time pressures in organi-zational life often lead even well-meaning, com-petent executives to neglect the social-emotionalaspects in their strategy implementation actions,and this explains, in part, the tall challenges ofmaking new strategy happen in firms.

ACKNOWLEDGEMENTS

Special thanks go to Special Issue Editor CraigFox and two anonymous reviewers who helped merefine the ideas in this paper. Appreciation alsogoes to: audiences at Harvard Business School,University of Michigan Ross School of Business,HEC Montreal, Organization Development andChange of the Academy of Management; mem-bers of my doctoral committee, Henry Mintzberg,Frances Westley, Ann Langley, and ChristianeDemers; Jan Jorgensen, Deborah Dougherty, andMichael Beer, who encouraged me in the earlydays of this research; my colleagues ChristophZott, Kevin Corley, Andrew Shipilov, and JavierGimeno. Finally, I gratefully acknowledge the sup-port of the INSEAD R&D committee, the INSEADAlumni Fund, and the Social Sciences HumanResearch Council of Canada.

REFERENCES

Argyris C, Schoen D. 1978. Organizational Learning .Addison-Wesley:Reading, MA.

Ashforth B, Harrison S, Corley KG. 2008. Identificationin organizations: an examination of four fundamentalquestions. Journal of Management 34(3): 325–374.

Ashforth BE, Humphrey RH. 1995. Emotion in theworkplace: a reappraisal. Human Relations 48:97–126.

Balogun J, Johnson G. 2004. Organizational restructuringand middle manager sensemaking. Academy ofManagement Journal 47: 523–549.

Barsade SG. 2002. The ripple effect: emotional contagionand its influence on group behavior. AdministrativeScience Quarterly 47: 644–675.

Bartunek JM. 1984. Changing interpretive schemes andorganizational restructuring: the example of a religiousorder. Administrative Science Quarterly 29: 355–372.

Biggart NW. 1977. The creative-destructive process oforganizational change: the case of the post office.Administrative Science Quarterly 22: 410–425.

Bornstein G, Budescue D, Zamir S. 1997. Cooperationin intergroup, two-person, and n-person games ofchicken. Journal of Conflict Resolution 41: 384–406.

Bourgeois LJ, Brodwin DR. 1984. Strategic implemen-tation: five approaches to an elusive phenomenon.Strategic Management Journal 5(3): 241–264.

Corley KG, Gioia DA. 2004. Identity ambiguity andchange in the wake of a corporate spin-off.Administrative Science Quarterly 49: 173–208.

Crossan MM, Berdrow I. 2003. Organizational learningand strategic renewal. Strategic Management Journal24(11): 1087–1105.

Daft RL, Macintosh NB. 1984. The nature and use offormal control systems for management control andstrategy implementation. Journal of Management 10:43–66.

Dutton JE, Dukerich JM. 1991. Keeping an eye onthe mirror: image and identity in organizationaladaptation. Academy of Management Journal 34:517–554.

Edmondson A. 1999. Psychological safety and learningbehavior in work teams. Administrative ScienceQuarterly 44: 350–383.

Eisenhardt KM. 1989. Assessing agency theory: anassessment and review. Academy of ManagementReview 14: 57–74.

Eisenhardt KM, Graebner ME. 2007. Theory buildingfrom cases: opportunities and challenges. Academy ofManagement Journal 50: 25–32.

Ellsworth PC, Scherer KR. 2003. Appraisal processesin emotion. In Handbook of Affective Sciences ,Davidson R, Goldsmith H, Scherer KR (eds). OxfordUniversity Press: New York; 572–595.

Fiol CM. 2002. Capitalizing on paradox: the role oflanguage in transforming organizational identities.Organization Science 13: 653–666.

Fiss PC, Zajac EJ. 2006. The symbolic managementof strategic change: sensegiving via framing anddecoupling. Academy of Management Journal 49(6):1173–1193.

Floyd SW, Wooldridge B. 1992. Middle managementinvolvement in strategy and its association withstrategic type: a research note. Strategic ManagementJournal, Summer Special Issue 13: 153–167.

Folkman S, Lazarus RS. 1988. The Ways of CopingQuestionnaire. Consulting Psychologists Press:PaloAlto, CA.

Fox S, Amichai-Hamburger Y. 2001. The power ofemotional appeals in promoting organizational changeprograms. Academy of Management Executive 15(4):84–94.

Frijda NH. 1986. The Emotions . Cambridge UniversityPress:Cambridge, U.K.

Frijda NH, Kuipers P, ter Schure E. 1989. Relationsamong emotion, appraisal, and emotional action readi-ness. Journal of Personality and Social Psychology 57:212–228.

Garcia SM, Ybarra O. 2007. People accounting: socialcategory-based choice. Journal of Experimental SocialPsychology 43: 802–809.

Garcia-Prieto P, Bellard E, Schneider SC. 2003. Expe-riencing diversity, conflict, and emotions in teams.Applied Psychology 52: 413–440.



George JM. 1995. Leader positive mood and groupperformance: the case of customer service. Journalof Applied Social Psychology 25: 778–794.

Gioia DA, Chittipeddi K. 1991. Sensemaking andsensegiving in strategic change initiation. StrategicManagement Journal 12(6): 433–448.

Gioia DA, Thomas JB. 1996. Identity, image, and issueinterpretation: sensemaking during strategic changein academia. Administrative Science Quarterly 41(3):370–403.

Gomez-Mejia LR, Palich LE. 1997. Cultural diversityand the performance of multinational firms. Journalof International Business Studies 28(2): 309–335.

Govindarajan V. 1988. A contingency approach tostrategy implementation at the business-unit level:integrating administrative mechanisms with strategy.Academy of Management Journal 31: 828–853.

Grandey AA. 2003. When ‘the show must go on:’ surfaceacting and deep acting as determinants of emotionalexhaustion and peer-rated service delivery. Academyof Management Journal 46: 86–96.

Guth WD, MacMillan IC. 1986. Strategy implementationversus middle management self-interest. StrategicManagement Journal 7(4): 313–327.

Hardin CD, Higgins ET. 1996. Shared reality: howsocial verification makes the subjective objective. InHandbook of Motivation and Cognition, Volume 3: TheInterpersonal Context , Sorrentino RM, Higgins ET(eds). Guilford Press: New York; 28–84.

Haslam SA, Reicher S. 2006. Stressing the group: socialidentity and the unfolding dynamics of responses tostress. Journal of Applied Psychology 91: 1037–1052.

Hochschild AR. 1983. The Managed Heart: Commercial-ization of Human Feeling . University of CaliforniaPress:Berkeley, CA.

Hogg MA, Terry DJ, White KM. 1995. A tale of twotheories: a critical comparison of identity theory withsocial identity theory. Social Psychology Quarterly 58:255–269.

Huy Q. 1999. Emotional capability, emotional intelli-gence, and radical change. Academy of ManagementReview 24: 325–345.

Huy Q. 2002. Emotional balancing of organizationalcontinuity and radical change: the contribution ofmiddle managers. Administrative Science Quarterly47: 31–69.

Izard CE, Ackerman BP. 2000. Motivational, organiza-tional, and regulatory functions of discrete emo-tions. In Handbook of Emotions (2nd edn). Lewis M,Haviland-Jones JM (eds). Guilford Press: New York;253–264.

Johnson G. 1988. Rethinking incrementalism. StrategicManagement Journal 9(1): 75–91.

Kreiner GE, Ashforth BE, Sluss DM. 2006. Identitydynamics in occupational dirty work: integratingsocial identity and system justification perspectives.Organization Science 17: 619–636.

Laumann EO, Pappi FU. 1976. Networks of CollectiveAction . Academic Press:New York.

Lazarus RS. 1993. From psychological stress to emotions:a history of changing outlooks. Annual ReviewPsychology 44: 1–21.

Lerner JS, Keltner D. 2001. Fear, anger, and risk. Journalof Personality and Social Psychology 81: 146–159.

Locke K. 1996. A funny thing happened! Themanagement of consumer emotions in serviceencounters. Organization Science 7: 40–81.

Mackie DM, Devos T, Smith ER. 2000. Intergroupemotions: explaining offensive action tendencies in anintergroup context. Journal of Personality and SocialPsychology 79: 602–616.

Maitlis S, Ozcelik H. 2004. Toxic decision processes: astudy of emotion and organizational decision making.Organization Science 15: 375–393.

Martin J, Knopoff K, Beckman C. 1998. An alternativeto bureaucratic impersonality and emotional labor:bounded emotionality at the Body Shop. Administra-tive Science Quarterly 43: 429–469.

Milgram S. 1992. The Individual in a Social World .McGraw-Hill:New York.

Miller D, Friesen PH. 1984. Organizations: A QuantumView . Prentice Hall:Englewood Cliffs, NJ.

Miller N, Brewer MB. 1986. Categorization effects oningroup and outgroup perception. In Prejudice,Discrimination, and Racism , Gaertner SL, Dovisio JF(eds). Academic Press: New York; 209–230.

Nutt PC. 1989. Selecting tactics to implement strategicplans. Strategic Management Journal 10(2): 145–161.

Pettigrew AM. 1985. Contextualist research: a naturalway to link theory and practice. In Doing Researchthat is Useful in Theory and Practice, Lawler EE (ed).Jossey-Bass: San Francisco, CA; 222–248.

Pratt MG. 2009. For the lack of a boilerplate: tipson writing up (and reviewing) qualitative research.Academy of Management Journal 52: 856–862.

Rangan S, Sengul M. 2009. The influence of macro struc-ture on the foreign market performance of transna-tional firms: the value of IGO connections, exportdependence, and immigration links. AdministrativeScience Quarterly 54(2): 229–267.

Robinson MD, Clore GL. 2002. Episodic and semanticknowledge in emotional self-report: evidence for twojudgment processes. Journal of Personality and SocialPsychology 83: 198–215.

Roseman IJ. 1991. Appraisal determinants of discreteemotions. Cognition and Emotion 5: 161–200.

Sanchez-Burks J. 2002. Protestant relational ideologyand (in)attention to relational cues in work settings.Journal of Personality and Social Psychology 83:919–929.

Sanchez-Burks J. 2005. Protestant relational ideology:the cognitive underpinnings and organizationalimplications of an American anomaly. Research inOrganizational Behavior 26: 265–305.

Sanchez-Burks J, Huy Q. 2009. Emotional aperture andstrategic change: the accurate recognition of collectiveemotions. Organization Science 20: 22–34.

Schein EH. 1992. Organizational Culture and Leadership(2nd edn). Jossey-Bass:San Francisco, CA.

Schmid C. 2001. The Politics of Language: Conflict,Identity, and Cultural Pluralism in ComparativePerspective. Oxford University Press:New York.

See KE. 2009. Reactions to decisions with uncertainconsequences: reliance on perceived fairness versus


1410 Q. N. Huy

predicted outcomes depends on knowledge. Journalof Personality and Social Psychology 96: 104–118.

Seo M-G, Barrett LF. 2007. Being emotional duringdecision making—good or bad? An empiricalinvestigation. Academy of Management Journal 50:923–940.

Simonin BL. 1997. The importance of collaborativeknow-how: an empirical test of the learningorganization. Academy of Management Journal 40(5):1150–1174.

Smith C. 1989. Dimensions of appraisal and physiolog-ical response in emotion. Journal of Personality andSocial Psychology 56: 339–353.

Smith CA, Ellsworth PC. 1985. Patterns of cognitiveappraisal in emotion. Journal of Personality and SocialPsychology 48: 813–838.

Smith ER, Seger CR, Mackie DA. 2007. Can emotionsbe truly group level? Evidence regarding fourconceptual criteria. Journal of Personality and SocialPsychology 93: 431–446.

Stahl GK, Voigt A. 2008. Do cultural differences matterin mergers and acquisitions? A tentative model andexamination. Organization Science 19(1): 160–176.

Stein NL, Trabasso T, Liwag M. 1993. The representa-tion and organization of emotional experience: unfold-ing the emotion episode. In Handbook of Emotions ,Lewis M, Haviland JM (eds). Guilford Press: NewYork; 279–300.

Suddaby R. 2006. From the editors: what grounded theoryis not. Academy of Management Journal 49: 633–642.

Sy T, Cote S, Saavedra R. 2005. The contagious leader:impact of the leader’s mood on the mood of groupmembers, group affective tone, and group processes.Journal of Applied Psychology 90: 295–305.

Tajfel H, Turner JC. 1979. An integrative theory ofintergroup conflict. In The Social Psychology ofIntergroup Relations , Austin WG, Worchel S (eds).Brooks/Cole: Monterey, CA; 33–47.

Thompson JD. 1967. Organizations in Action. McGraw-Hill:New York.

Tichy NM, Sherman S. 1994. Control Your Destiny orSomeone Else Will . Harper Business:New York.

Tiedens LZ, Linton S. 2001. Judgment under emotionaluncertainty: the effects of specific emotions andtheir associated certainty appraisals on informationprocessing. Working paper, Stanford University.

Turner JC. 1991. Social Influence. Wadsworth Publish-ing:Belmont, CA.

Turner JC, Hogg MA, Oakes PJ, Reicher SD, WetherellMS. 1987. Rediscovering the Social Group: A Self-Categorization Perspective. Basil Blackwell:Oxford,U.K.

Uyterhoeven H. 1989. General managers in the middle.Harvard Business Review 5: 136–145.

Vaara E, Tienari J, Piekkari R, Santti R. 2005. Languageand the circuits of power in a merging multinationalcorporation. Journal of Management Studies 42:595–623.

van Zomeren M, Postmes T, Spears R. 2008. Toward anintegrative social identity model of collective action:a quantitative research synthesis of three social-psychological perspectives. Psychological Bulletin144(4): 504–535.

van Zomeren M, Spears R, Fischer AH. 2004. Put yourmoney where your mouth is! Explaining collectiveaction tendencies through group-based anger andgroup efficacy. Journal of Personality and SocialPsychology 87: 649–664.

Westley FR. 1990. Middle managers and strategy:microdynamics of inclusion. Strategic ManagementJournal 11(5): 337–351.

Yin RK. 1994. Case Study Research (2nd edn).Sage:Thousand Oaks, CA.

Zott C, Huy Q. 2007. How entrepreneurs use symbolicmanagement to acquire resources. AdministrativeScience Quarterly 52: 70–105.


Documents

HOW MIDDLE MANAGERS’ GROUP-FOCUS EMOTIONS AND … › quy-huy › documents › SMJ 2011... · 2015-11-27 · Middle Managers Group-Focus Emotions and Strategy Implementation 1389