CEO Personality Nadkarni y Herrman (2010)

7/22/2019 CEO Personality Nadkarni y Herrman (2010)

1/25

CEO PERSONALITY, STRATEGIC FLEXIBILITY, AND FIRMPERFORMANCE: THE CASE OF THE INDIAN BUSINESS

PROCESS OUTSOURCING INDUSTRY

SUCHETA NADKARNI

Drexel University

POL HERRMANNIowa State University

We examine the relationships between CEO personality, strategic flexibility (ability toadapt quickly to environmental changes), and firm performance, using a sample of 195small and medium-sized firms from the Indian business process outsourcing industry.We hypothesize that strategic flexibility mediates the relationships between CEOpersonality and firm performance. Our results extend previous research by not onlyhighlighting the importance of CEO personality in driving strategic flexibility, but alsoindicating how each facet of CEO personality either enhances or inhibits strategicflexibility.

With increasingly intense competition, shrinkingproduct cycles, accelerated technological break-throughs, and progressively greater globalization,the business arena may best be described as beingin a chronic state of flux, with continual variationin its external environment. Given such ever-changing environmental conditions, a firms abilityto change direction quickly and to reconfigure stra-tegically is crucial to its success in achieving sus-tainable competitive advantage (Hitt, Keats, & De-Marie, 1998). In other words, firms need to embracestrategic flexibility(Hitt et al., 1998; Johnson, Lee,Saini, & Grohmann, 2003). Ample empirical evi-dence supports the contention that strategic flexi-

bility drives firm performance (Grewal & Tansuhaj,2001; Nadkarni & Narayanan, 2007; Worren,Moore, Cardona, 2002). It is therefore not surpris-ing that the academic and practitioner literature instrategic management is increasingly recognizingstrategic flexibility as an important research area.

Nevertheless, several gaps remain in scholarsunderstanding of how firms embrace strategicflexibility. One particularly prominent gap re-

lates to the role of CEOs in fostering strategicflexibility. A great deal of the research that hasexamined the influence of resource, product, andalliance network structures on strategic flexibility

(Sanchez, 1995; Worren et al., 2002; Young-Ybarra& Wiersema, 1999) has ignored the role of CEOs indeveloping strategic flexibility. This gap is espe-cially notable because the strategic choice (Child,1972) and upper echelons (Hambrick & Mason,1984) perspectives have highlighted the impor-tance of top managers, especially CEOs, in drivingstrategic changes in firms (Rajagopalan & Spreitzer,1997). The CEO has been characterized as a firmschief cognizer and decision maker (Calori, Johnson,

& Sarnin, 1994). Hambrick and Mason (1984) ar-gued that firm strategies reflect the characteristicsof its powerful actors, among whom the CEO isprominent. Moreover, empirical evidence has sug-gested that characteristics of CEOs affect strategicdecision processes (Peterson, Smith, Martorana, &Owens, 2003) and strategic actions (Carpenter,Sanders, & Gregersen, 2001; Miller & Toulouse,1986; Nadkarni & Narayanan, 2007) that have im-plications for firm performance. However, thesestudies have examined the influence of CEO per-sonality on firm performance without paying ade-quate attention to the mechanisms that underliethis relationship (Peterson et al., 2003). The

bounded rationality (Simon, 1991) and managerialcognition (Weick, 1995) literatures have suggestedcognitive filtering mechanisms that may explainhow attributes of CEOs dispose them toward spe-cific strategic behaviors with implications for firmperformance. Examining these underlying mecha-nisms by integrating insights from the literatures onstrategic flexibility, managerial cognition, and CEOattributes provided the primary motivation for thecurrent study.

We would like to thank professors Deepak Datta andMary Uhl-Bien for their helpful comments on earlierversions of this article. We especially thank AssociateEditor Gerard Sanders and the three anonymous review-ers for their excellent and developmental feedback,which helped us immensely in improving the paper.

Academy of Management Journal

2010, Vol. 53, No. 5, 10501073.

1050Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holders express

written permission. Users may print, download or email articles for individual use only.


2/25

We address this theoretical gap by examining therelationships among CEO personality, strategicflexibility, and firm performance. Drawing on theupper echelons (Finkelstein & Hambrick, 1996;Hambrick & Mason, 1984) and CEO psychology(Chatterjee & Hambrick, 2007; Hiller & Hambrick,2005) literatures, we theorize that psychological

attributes of CEOs serve as lenses through whichCEOs subjectively view strategic situations and de-cide on appropriate responses, by shaping theirfields of vision(how CEOs acquire and disseminateinformation), their selective perception (which in-formation cues from their fields of vision CEOsattend to and which cues they choose to ignore),and their interpretation of perceived cues (howCEOs attach meaning to noticed cues and how theyevaluate strategic options). These filtering mecha-nisms that form the basis for a CEOs strategicchoices either enhance or inhibit strategic flexibil-ity in a dynamic industry context. We further the-orize that strategic flexibility influences firm per-formance by promoting creativity, innovation, andimproved competitive capability (Hitt et al., 1998;

Johnson et al., 2003). Specifically, we propose thatstrategic flexibility mediates the relationship be-tween CEOs personality attributes and firm perfor-mance. We theorize and empirically test these re-lationships for a dynamic industry context becauseresearch has suggested that the influence of strate-gic flexibility on firm performance is likely to bestronger in dynamic industries than in stable ones(Hitt et al., 1998; Johnson et al., 2003; Nadkarni &

Narayanan, 2007).We tested our model in data on the CEOs of small

and medium-sized enterprises (SMEs) from the off-shore business process outsourcing industry in In-dia. Offshore business process outsourcing is thetransfer of the operational ownership of one ormore of a firms processes to an external providerfrom another country that then manages the pro-cesses according to predetermined metrics (Ghosh& Scott, 2005). It is becoming a widespread strat-egy, and India is a dominant service provider inthis field, accounting for 75 percent of offshore

delivery value (Neale, 2004). Indian business pro-cess outsourcing is a fast growing and dynamicindustry characterized by low barriers to entry, rap-idly changing and unpredictable process technolo-gies, ever-changing client demands, shifting globalcompetition, and constant client pressure to im-prove value and delivery speed (Ramachandran &Voleti, 2004; Tapper, 2004). Anecdotal evidencesuggests that strategic flexibility is essential for suc-cess in this industry (Mehta, Armenakis, Mehta, &Irani, 2006). Our results extend previous research

by highlighting the role of CEO personality in de-

veloping strategic flexibility and by demonstratinghow specific facets of CEO personality influencefirm performance by either enhancing or inhibitingstrategic flexibility in a fast-changing and dynamicindustry context.

THEORETICAL BACKGROUND

Strategic Flexibility

Strategy scholars have defined strategic flexibil-ityas a firms ability to precipitate strategic changes(Evans, 1991; Harrigan, 1985). Aaker and Mas-carenhas (1984) defined it as the ability to adapt tosubstantial, uncertain, and rapidly occurring envi-ronmental changes that meaningfully impact firmperformance. Thus, strategic flexibility reflects afirms ability to respond continuously to unantici-pated changes and to adjust to unexpected conse-quences of predictable changes (Lei, Hitt, & Gold-

har, 1996).Most studies of strategic flexibility have focused

on technology (Evans, 1991; Sanchez, 1995; Wor-ren et al., 2002), resources (Harrigan, 1985; Young-Ybarra & Wiersema, 1999), and network structures(Young-Ybarra & Wiersema, 1999) as antecedents.For example, Sanchez (1995) found that productand process platform architectures drove strategicflexibility, whereas Evans (1991) focused on theeffects of technological maneuvers. Asset specific-ity (Young-Ybarra & Wiersema, 1999) and immobil-ity of resources (Harrigan, 1985) have also been

identified as antecedents of strategic flexibility.These studies have ignored the influence of CEOson strategic flexibility. We propose that personalityattributes of CEOs influence strategic flexibility.

Overview of the Literature on CEO Psychology

Strategy research has suggested that a firms CEO,as an important member of the firms dominantcoalition, has a profound impact on the strategicdirection and performance of the firm (Hambrick &Mason, 1984; Peterson et al., 2003). Hambrick

(1994) criticized studies that treat a CEO as justanother member of a top management team (TMT),noting that everyday observation and empirical ev-idence indicate that the CEO has a disproportion-ate, sometimes dominating, influence on his or herfirm. Finkelstein and Hambrick (1996) asserted thatnot only does the CEO have the overall responsi-

bility for the firms management but also that theCEOs characteristics are of serious consequence tothe firm.

Researchers studying CEOs have often used de-mographic characteristics as proxies for deeper

2010 1051Nadkarni and Herrmann


3/25

psychological constructs (Carpenter, Geletkanycz,& Sanders, 2004). Demographic variables such asage, education, and experience allow researchers toeffectively capture characteristics such as back-ground and expertise, which are relevant to howCEOs make decisions (Hambrick & Mason, 1984).However, the use of demographic characteristics as

proxies for CEOs psychological traits leaves re-searchers at a loss as to the real psychological at-tributes that drive CEO behavior (Carpenter et al.,2004), undermines the robustness of theories re-garding the links between psychological character-istics of CEOs and firm outcomes, and increases thelikelihood of incorrect interpretation of results(Lawrence, 1997). Lawrence (1997) referred tothese limitations as a black box. To overcome them,recent studies have focused on CEO psychology,describing how CEOs broadly evaluate themselvesand their relationships to their environmentsacross situations (Hiller & Hambrick, 2005). Theunderlying premise of this research is that CEOsconfront so many stimuli, laden with so much am-

biguity, complexity, and contradiction, that theirpersonalities greatly enter into how they distill andprocess this information. Psychological attributesof CEOs, by filtering how CEOs construe the realityof strategic situations and evaluate strategic re-sponse options, dispose the CEOs toward certainchoices (Finkelstein & Hambrick, 1996).

There is empirical support for the contention thatCEO personality attributes influence their strategicchoices, which in turn influence firm performance.

Miller and Toulouse (1986) found that CEOs withinternal loci of control deployed product innova-tion strategies, whereas CEOs with high needs forachievement chose broad market strategies. Hay-ward and Hambrick (1997) found that CEO hubris,manifested as exaggerated pride or self-confidence,was positively related to paying acquisition premi-ums and negatively related to firm performance.Chatterjee and Hambrick (2007) suggested that nar-cissistic CEOs chose bold strategies (e.g., large ac-quisitions) that attract attention, resulting in bigwins or big losses.

A Five-Factor Model of CEO Personality

We focus on the effect of personality variables ascaptured by the five-factor model (McCrae &Costa, 1987). Our choice of this model was basedon recent calls to use comprehensive and validpsychological frameworks to investigate the rela-tionships between CEOs personality attributes andfirm performance (Cannella & Monroe, 1997; Hiller& Hambrick, 2005). The five-factor model, whichrepresents current orthodoxy in personality assess-

ment, provides a robust, comprehensive way ofunderstanding fundamental personality differences(Peterson et al., 2003). Although opinion is not yetunanimous, there is increasing consensus amongresearchers that the traits identified in the five-factor model encapsulate many important aspectsof personality (Judge, Bono, Ilies, & Gehardt, 2002;

McCrae & Costa, 1997). Strategy researchers havealso underscored the importance of this model inexplaining behaviors of top managers such as CEOs(Cannella & Monroe, 1997), and recent empiricalevidence has underscored its relevance to strategicdecision making as well (Peterson et al., 2003).

The five factors are broad personality constructs,each capturing a unique set of psychological traits(Boudreau, Boswell, Judge, & Bretz, 2001). Consci-entiousness indicates achievement and depend-ability.Emotional stabilityis the ability to adapt todiverse situations and to cope with stress. Agree-ableness is the tendency to be altruistic and com-pliant. Extraversion represents sociability and ex-pressiveness. Openness to experience representsthe tendency to be creative, imaginative, percep-tive, and thoughtful.

THEORY DEVELOPMENT AND HYPOTHESES

CEO Personality, Strategic Flexibility, andFirm Performance

The upper echelons (Finkelstein & Hambrick,1996; Hambrick & Mason, 1984) and CEO psychol-

ogy (Hiller & Hambrick, 2005) literatures suggestthat psychological attributes of CEOs influencetheir strategic choices through a three-stage filter-ing processdefining a field of vision, selectiveperception, and interpretation. This filtering pro-cess is considered to be central to developing stra-tegic flexibility (Johnson et al., 2003; Nadkarni &Narayanan, 2007; Shimizu & Hitt, 2004).

Psychological attributes determine how in-tensely CEOs search for information, how muchinformation they scan, how they learn about exter-nal environmental and internal organizational

events or trends, and which sources they rely on toobtain and disseminate information (Hambrick,1982; Miller & Toulouse, 1986). These activitiesdefine a CEOs focus of attention or field of vision,which serves as a filter between an objective stra-tegic situation and the subjective reality of the sit-uation construed by the CEO, wrote Finkelsteinand Hambrick (1996). They proposed, for example,that a CEO with an internal locus of control willdevote more effort to environmental scanning byusing a wider array of sources than an executivewith an external locus of control will use. Conse-

1052 OctoberAcademy of Management Journal


4/25

quently, internally focused CEOs develop broaderfields of vision than externally focused CEOs. Nad-karni and Narayanan (2007) proposed that such a

broad field of vision fosters strategic flexibility byenabling a firm to develop a comprehensive aware-ness of new opportunities and new resources andalso helps the firm to change its competitive pos-

ture quickly, by promoting better understanding ofcontinuously shifting competitor moves. Johnsonet al. (2003) also stressed that the panoramic sur-veillance made possible by a broad field of visionimproves the market-sensing capability that iscentral to strategic flexibility.

Second, research has suggested that top manag-ers selectively perceive only a small fraction of thestimuli within their fields of vision (Starbuck &Milliken, 1988). Which stimuli CEOs attend to andwhich they ignore is tied to their psychologicalattributes (Finkelstein & Hambrick, 1996), such asopenness to change (Datta, Rajagopalan, & Zhang,2003) and need for achievement (Miller & Droge,1986). Shimizu and Hitt (2004) stressed that astrong selective perception bias is a major barrier todeveloping strategic flexibility because it preventsstrategic decision makers from being sensitive toimportant new information and makes them com-placent. Johnson et al. (2003) also proposed thatselective perception bias can filter out importantmarket events and inhibit the responsive capabilityof a firm, which is also central to strategicflexibility.

Interpreting or attaching meaning to perceived

stimuli is the final step in the filtering process. Itconsists of understanding, explaining, extrapolat-ing, and predicting the effect of strategic stimuli(Starbuck & Milliken, 1988). Such interpretationforms the basis for the evaluation and choice ofstrategic options. Psychological attributes such asrisk propensity and need for control may influencewhether CEOs interpret specific environmentalchanges as threats or as opportunities and whichstrategic responses they prefer (Finkelstein & Ham-

brick, 1996). Shimizu and Hitt (2004) underscoredthe importance of timely and effective interpreta-

tion in developing strategic flexibility. For exam-ple, strategic decision makers may interpret earlynegative results of a strategy to be a sign of incorrectimplementation or insufficient time rather than asign of the ineffectiveness of the strategy. Suchmisinterpretation prompts firms to invest more re-sources in outdated and obsolete strategies ratherthan recognize the need to abandon them. Inertiaand barriers to strategic flexibility thus arise(Shimizu & Hitt, 2004).

We integrated the literatures on upper echelonsand strategic flexibility to develop hypotheses for

each facet of the five-factor model. We outline howeach facet is likely to influence the filtering viafield of vision, selective perception, and interpre-tation that is central to developing strategicflexibility.

CEO Personality and Strategic Flexibility

Conscientiousness. Conscientiousness reflectsthe degree to which someone showsdependabilityand anachievement orientation(Judge et al., 2002;McCrae & Costa, 1997). Dependability is a concernfor legalism or commitment to established rules(Peterson et al., 2003). Individuals with high de-pendability avoid taking actions that deviate signif-icantly from their past experience. An achievementorientation represents a need for control and a needto receive concrete feedback on actions (Miller &Droge, 1986). High achievers feel a strong need totake responsibility for doing things immediately.

Because of their concern for legalism, conscien-tious CEOs are likely to rely strongly on depend-able, tried-and-true strategies. Over time, as CEOsrely almost exclusively on known strategies andselectively ignore new and unique strategies thatchallenge their existing assumptions, they arelikely to develop narrow fields of vision and aselective perception bias that predisposes them toignore environmental stimuli that do not matchexisting assumptions (Bogner & Barr, 2000; Kiesler& Sproull, 1982). Such a narrowed field of visionand strong selective perception bias create strong

barriers to strategic flexibility by inhibiting themarket-sensing capability (Johnson et al., 2003).When CEOs fail to see important environmentalstimuli that do not fit their narrow visions, theywill be unable to respond to critical environmentalchanges. This will inhibit their ability to quicklyinitiate strategic responses (Nadkarni & Narayanan,2007).

Achievement-oriented CEOs also feel the need topersonally take control and assume responsibilityfor strategic activities. Miller and Toulouse (1986)and Miller and Droge (1986) found that CEOs with

high needs for achievement tend to hold most ofthe power in their own hands and to closely mon-itor and control employee activities in their firms.Such closely controlled and highly structured de-cision making is likely to deny creative employeesthe autonomy and freedom to question existing as-sumptions, create new interpretations, and shareinformation freely in a firm, resulting in a narrowfield of vision (Choo, 1998). Lack of rigorous debateand discussion of strategic issues among employeeswith varied backgrounds also creates the potentialfor selective perception and interpretation biases



5/25

(Lant, Milliken, & Batra, 1992), which inhibit stra-tegic flexibility by undermining ability to sensenew and unfamiliar information in a timely manner(Johnson et al., 2003; Shimizu & Hitt, 2004) as wellas ability to initiate responsive actions quicklythrough efficient resource deployment (Hitt et al.,1998; Nadkarni & Narayanan, 2007).

Conscientious individuals have a strong need toreduce uncertainty and to receive specific feedbackon their performance (Judge et al., 2002). For con-scientious CEOs, performance feedback and long-range planning are central to making strategicchoices (Miller & Droge, 1986; Miller & Toulouse,1986). Researchers have described such a perfor-mance-driven approach to strategy formulation as acompetence trap, because it creates a strong se-lective perception bias by disposing strategic deci-sion makers to ignore new and different environ-mental information unless significant performancedeclines occur (Bogner & Barr, 2000; Brown &Eisenhardt, 1997; Johnson et al., 2003). Thus, con-scientious CEOs may not attend to ambiguous anduncertain cues until performance declines alertthem to the need for strategic change. Waiting forperformance declines to signal the need for devel-oping new strategic thinking can create delays instrategic decision making and impede responsivecapability (Eisenhardt & Martin, 2000; Shimizu &Hitt, 2004). To develop efficient responsive capa-

bility, strategic decision makers need to engage ininterpretation and search activities that are intui-tive and exploratory rather than feedback-oriented

(Bogner & Barr, 2000; Daft & Weick, 1984). Recentevidence from organizational behavior supportsthis negative relationship between conscientious-ness and the ability to adapt to changing contexts.In their experimental study, Lepine, Colquitt, andErez (2001) found that participants with low con-scientiousness adapted better to changing taskcontexts.

Hypothesis 1. CEO conscientiousness is nega-tively related to strategic flexibility.

Emotional stability. Emotional stability reflects

a capacity foremotional adjustmentand self-confi-dence. Emotional adjustment is the ability of indi-viduals to adjust their emotional states to variedsituational demands and to remain calm and bal-anced in stressful situations (McCrae & Costa,1997). Emotional stability is considered a strongpredictor of a persons adaptability to unpredict-able and changing situations (Peterson et al., 1993).Research suggests that the emotional stability of aleader is more relevant to decision making inchanging and unpredictable situations than in sta-

ble ones. For example, De Hoogh, Den Hartog, and

Koopman (2005) found that emotional stability pre-dicted leader effectiveness for dynamic but not forstable tasks. Studies have shown that emotionalstability is also strongly associated with internallocus of control and leader attributes associatedwith this construct (Judge et al., 2002). Given thesalience of emotional stability in decisions in dy-

namic situations, we hypothesize that CEOs emo-tional stability promotes strategic flexibility.Emotionally stable managers remain calm and

provide focus in dynamic situations, shift focus toinitiate appropriate actions to deal with unpredict-able situations, and act decisively in crises (Peter-son et al., 2003). Emotionally stable leaders create asafe atmosphere for employees by reducing theiranxiety in difficult situations and by providing en-couragement in cases of failure (Edmondson, 1999).This outline suggests that emotionally stable CEOsfeel less threatened by new and unpredictable stim-uli and encourage employees to experiment withnew interpretations of these stimuli. The adaptabil-ity of emotionally stable CEOs reduces their hesi-tance to change strategies and enables them toquickly generate appropriate responses to thesechanges. Such a balanced and adaptive approachallows a CEO to process adverse and ambiguousinformation objectively and rationally, and thismanner of responding is likely to evoke a broadfield of vision and to reduce selective perceptionand interpretation biases. Consequently, emotion-ally stable CEOs are likely to improve their sensingand responsive capabilities, which are central to

developing strategic flexibility (Johnson et al.,2003; Shimizu & Hitt, 2004).

Studies have shown that because emotionallystable leaders have high self-confidence, they arenot afraid to challenge the status quo. Overcomingorganizational inertia, an important barrier to stra-tegic flexibility (Shimizu & Hitt, 2004), requireschallenging the status quo and taking risks, both ofwhich require the high degree of self-confidencetypical of emotionally stable leaders (House &Howell, 1992; Kirkpatrick & Locke, 1991). Thus,the confidence and decisiveness of an emotionally

stable CEO may promote strategic flexibility by re-moving barriers such as organizational inertia andfiltering biases.

Hypothesis 2. CEO emotional stability is posi-tively related to strategic flexibility.

Agreeableness. Agreeableness represents thetendency to bealtruistic(empathetic, kind, cooper-ative, trusting, and gentle) and compliant(modest,having a values affiliation, and conflict avoiding)(Bono & Judge, 2004). The relationship betweenleader agreeableness and the ability to bring about



6/25

change is ambiguous because of two underlying,opposing mechanisms. On the one hand, agreeable-ness (altruism and compliance) fosters a culture ofcreativity and risk taking based on cooperative,open, and trust-based relationships with employ-ees (Judge & Bono, 2000). On the other hand, ex-cessive agreeableness can also give rise to passivity,

in which leaders act modest, focus more on whatemployees think of them than on accomplish-ments, and avoid conflicts at all costs. These twosets of mechanisms evoke leader behaviors (Lan-gan-Fox, Cooper, & Klimoski, 2007) that may eitherenable or inhibit adaptability and innovation(LePine & Van Dyne, 2001). We propose that me-dium levels of agreeableness allow CEOs to opti-mally balance these opposing mechanisms so as tomaximize strategic flexibility, whereas very highlevels of agreeableness induce passivity and com-pliance and very low levels undermine employeecreativity and risk taking, inhibiting strategicflexibility.

Disagreeable CEOs promote a climate of compe-tition and fear (Peterson et al., 2003) that is likelyto promote compliance rather than independentthinking. Intimidated by disagreeable CEOs, em-ployees are hesitant to bring to their attention in-formation that may challenge the CEOs personal

beliefs (Peterson et al., 2003), and this hesitancynarrows the CEOs fields of vision. Moreover, dis-agreeable CEOs may be skeptical of and ignore thestrategic alternatives suggested by other managersand employees. This behavior may create strong

perceptual and interpretation biases (Lant et al.,1992), which inhibit strategic flexibility (Nadkarni& Narayanan, 2007; Shimizu & Hitt, 2004).

Highly agreeable leaders pay special attention toneglected groups in their firms, treat each em-ployee as an individual, and focus on employeeempowerment, which fosters free and comprehen-sive exchange of information between diverse em-ployees (Bono & Judge, 2004). Such comprehensiveinformation exchange is likely to broaden a CEOsfield of vision (Lant et al., 1992). However, thestrongly altruistic tendency of highly agreeable

leaders can promote passivity and compliance andshift focus away from achievement of importanttask goals. For example, Langan-Fox et al. (2007)contended that because highly agreeable individu-als value and strive for cooperation and harmony,they may avoid engaging in certain functional task-focused behaviors when their behaviors have thepotential to upset other individuals with whomthey work, which is likely to inhibit decision effec-tiveness. Similarly, Lepine and Van Dyne (2001)found that high agreeableness inhibited voice be-havior, which is defined as the extent to which an

individual speaks up with constructive suggestionsfor change. They found that high levels of agree-ableness were detrimental in situations of innova-tion and adaptability because voice behavior wassuppressed. This evidence suggests that a strongneed for affiliation and concern about what othersthink of them may suppress the voice behaviors of

CEOs and prompt them to surrender their views insituations of conflict rather than engage in thestrong influencing tactics needed to foster respon-sive capabilities. Thus, the perceptions and inter-pretations of highly agreeable CEOs may be drivenprimarily by their need for affiliation and socialacceptance rather than by a decision focus based onobjective information, a focus that is likely to createstrong selective perception and interpretation bi-ases that create barriers to developing strategic flex-ibility (Shimizu & Hitt, 2004).

CEOs with medium levels of agreeableness maymaximize strategic flexibility by balancing em-ployee concern and empowerment with the strongand assertive voice, rhetoric, and assertiveness thatare needed to build a culture of change. Buildingstrategic change capability involves assessing hid-den assumptions, unlearning old behaviors, andovercoming major obstacles (Senge, 1990; Shimizu& Hitt, 2004). Moderately agreeable CEOs may

broaden their fields of vision by empowering em-ployees to generate new and controversial ideasthat challenge existing assumptions and behaviors.At the same time, CEOs can exercise assertivenessin the situations of conflict among departmental

and operational managers about possible organiza-tional changes that typically occur in the processof building the capability for strategic change(Burgelman, 1984). These conflicts tend to slowdown and freeze the capability-building process,unless CEOs and other strategic leaders activelyintervene with strong rhetoric and effective persua-sion (Elenkov, Judge, & Wright, 2005). CEO asser-tiveness can reduce the perceptual and interpreta-tion biases resulting from passivity and excessiveneed for affiliation. Therefore, we expect CEOswith medium levels of agreeableness to maximize

strategic flexibility.Hypothesis 3. CEO agreeableness has aninverted-U relationship with strategicflexibility.

Extraversion.Extraversion is associated with so-ciability and expressiveness (Judge et al., 2002).Extraverted leaders tend to take the initiative insocial settings, to introduce people to each otherand to be socially engaging by being humorous,introducing topics of discussion, and stimulatingsocial interactions (House & Howell, 1992). There-



7/25

fore, extraverted leaders build broad and diversenetworks of social relationships. Extraverted lead-ers are expressive and articulate individuals whopersuade, influence, and organize others (Bono &

Judge, 2004).The sociability of extraverted CEOs allows them

to mobilize others and to develop extensive social

interactions both internally (within their firms) andexternally (outside the firms). CEOs networks ofcontacts are central determinants of their fields ofvision (McDonald & Westphal, 2003). CEOs usethese networks to both receive and disseminateinformation (Kotter, 1982). Extensive social inter-actions result in comprehensive information gath-ering, support interpretation of new information(Kraatz, 1998), and promote its speedy transmis-sion (Davis & Greve, 1997). McDonald, Khanna,and Westphal suggested that CEOs who developextensive advice networks (2008: 453) are ex-posed to alternative and novel points of view; thisexposure enhances CEOs ability to quickly iden-tify the strategic challenges facing their companiesand develop high-quality solutions to them. Use of

broad networks for information acquisition anddissemination allows for intensive discussion andvalidation of new information, reducing selectiveperception and interpretation biases (McDonald &Westphal, 2003). Reduction in these biases reduces

barriers and promotes strategic flexibility (Nad-karni & Narayanan, 2007; Shimizu & Hitt, 2004).

Developing the ability to quickly adapt to envi-ronmental changes requires creation of new ideas

that may deviate from past strategies (Johnson etal., 2003), and the newness of such strategies mayitself create resistance among employees (Kirk-patrick & Locke, 1991), which can create inertiaand barriers to strategic flexibility (Hitt et al., 1998;Shimizu & Hitt, 2004). Extravert CEOs can effec-tively remove such resistance and promote rapidimplementation of new strategies through their ex-ceptional expressive skills and their ability to takethe initiative and persuade and influence people soas to promote strategic flexibility.

Hypothesis 4. CEO extraversion is positively

related to strategic flexibility.

Openness to experience.People who are open tonew experiences are intellectually curious, open toa wide range of stimuli, value unusual thoughtprocesses, and often seen as thoughtful and cre-ative (McCrae & Costa, 1987). Open individualshave a strong need for change and are highly capa-

ble of understanding and adapting to others per-spectives (Costa & McCrae, 1988). Leaders who areopen to new experiences actively seek excitementand risks (Judge et al., 2002). This need for change

and risk taking can promote behaviors that maydisrupt the existing product and resource advan-tages of stable firms (Nadkarni & Narayanan, 2007).However, CEOs openness to new experiences iscentral to promoting strategic adaptation in dy-namic environments (Datta et al., 2003).

Developing the capability to precipitate strategic

change requires that strategic leaders (CEOs) under-stand and adapt to multiple perspectives and thatthey be open to and accepting of strategic change(Black & Boal, 1996). Because of their broad inter-ests, divergent thinking, and receptiveness to awide range of stimuli, CEOs with high openness toexperience are likely to develop broad fields ofvision by considering multiple strategic perspec-tives. Open CEOs can quickly and effectively noticeand interpret new and diverse environmental infor-mation that does not fit the existing mind-set andare likely to consider a wide range of strategic al-ternatives, including those that deviate greatly fromexisting strategies. Thus, open CEOs are likely tominimize selective perception and interpretation

biases, which inhibit strategic flexibility (Johnsonet al., 2003; Nadkarni & Narayanan, 2007; Shimizu& Hitt, 2004). In contrast, executives who are averseto new experiences are likely to possess relativelyrestricted fields of vision within which to seek(Cyert & March, 1963) and evaluate alternatives(Finkelstein & Hambrick, 1996). Such CEOs, overtime, develop habits, establish routine informationsources, and rely mostly on past experience (Dattaet al., 2003). These biases, which may lead to ig-

noring important new stimuli that do not fit theirfields of vision, interpreting new stimuli inappro-priately, and avoiding effective strategic responseoptions that deviate from past strategies (Kiesler &Sproull, 1982), create strong barriers to strategicflexibility (Shimizu & Hitt, 2004).

Hypothesis 5. CEO openness to experience ispositively related to strategic flexibility.

Strategic Flexibility and Firm Performance

We propose that CEO personality influences firmperformance by fostering strategic flexibility. Inother words, strategic flexibility mediates the rela-tionship between CEO personality and firm perfor-mance. Ample theoretical and empirical evidencesupports the relationship between strategic flexibil-ity and firm performance. In todays businessenvironments, products, market, and competitive

boundaries are in a state of continuous flux (Evans,1991; Johnson et al., 2003; Nadkarni & Narayanan,2007). To compete effectively in such intenselycompetitive and technologically changing environ-



8/25

ments, firms need to develop strategic flexibility byrecalibrating their strategies and refocusing re-sources on successive decision points, often withdifferent rules of engagement (Bahrami & Evans,1989). Stability may lock company resources intooutdated products and processes, adversely affect-ing performance (Nerkar & Roberts, 2004). Hitt et al.

(1998) argued that in todays competitive land-scape, characterized by increasing strategic discon-tinuities, disequilibrium, hypercompetition, inno-vation, and continuous learning, firms successdepends on their ability to respond quickly tochanging competitive conditions (strategic flexibil-ity). Sanchez, Heene, and Thomas (1996) statedthat strategic flexibility allows for the attainment ofhigh performance and the ability to take advantageof firm opportunities. Nadkarni and Narayanan(2007) and Grewal and Tansuhaj (2001) found em-pirical support for the positive relationship be-tween strategic flexibility and firm performance indynamic environments.

Hypothesis 6. Strategic flexibility is positivelyrelated to firm performance.

Hypothesis 7. Strategic flexibility mediates therelationship between CEO personality and firmperformance.

METHODS

Setting

Four factors guided our choice of the Indian busi-ness process outsourcing industry as our researchsetting. First, this industry is gaining increasingimportance in both the academic and practitionerliterature in management. Offshore business pro-cess outsourcing has become a widespread strategy,with a projected annual growth rate of 60 percent(Tapper, 2004); in 2004, over 40 percent ofFortune500 companies were estimated to have outsourcedactivities offshore to enjoy cost and time advan-tages (Mehta et al., 2006). Offshore firms provide avariety of services, including customer support,

back-office transaction processing, informationtechnology and software operations, finance andaccounting services, and human resource services(Nag, 2004). Paralleling the popularity of businessprocess outsourcing is increasing academic recog-nition of the importance of examining this industry(Dibbern, Winkler, & Heinzl, 2008; Levina & Vaast,2008). However, few studies have empirically ex-amined strategy issues in the industry. India is theleader in business process outsourcing services,controlling 75 percent of offshore delivery value(Neale, 2004). The scarcity of empirical studies us-

ing samples of firms from this industry and thedominance of Indian firms in it were the primarymotivations for our choice of Indian business pro-cess outsourcing as setting.

Second, along with spectacular growth rates, theIndian business process outsourcing industry has

been experiencing many competitive shifts as a

result of low barriers to entry and an influx of newcompetitors, both domestically and globally (e.g.,from countries such as China and the Philippines).The new and different rules of engagement de-ployed by the new entrants have created majorshifts in competitive spaces in the offshore busi-ness process outsourcing markets in India (Ram-achandran & Voleti, 2004). Other reasons for thedynamism include the numerous and unpredict-able changes in communication and process tech-nologies, constantly shifting client needs, and rad-ical changes in client businesses (Mehta et al.,2006). Steady improvement of products and ser-vices is no longer sufficient for surviving in theglobal market. To cope with the rapidity of change,

business process outsourcing firms need to developnew areas of technical and business domain exper-tise (Nag, 2004), improve delivery speed and valueto clients, and find radically new ways of develop-ing new service products (Ramachandran & Voleti,2004). Their survey of Indian business process out-sourcing managers led Mehta et al. (2006) to con-clude that to successfully meet the challenges inthe industry, firms must encourage employees tothink outside the box, develop the ability to

adapt to change, and foster a learning culture.Thus, strategic flexibility is central to survival, notto mention success, in this Indian industry.

Third, the majority of the firms in the Indianbusiness process outsourcing industry are SMEsfounded by entrepreneurs; 92 percent of the CEOsin our sample founded their companies. Finkel-stein and Hambrick (1996) argued that because ofthe small size of operations and the dual roles ofCEOs as both owners and managers, CEOs in suchSMEs enjoy considerably more power in strategyformulation and implementation than do their

counterparts in large firms, where ownership isseparated from management and the large size ofoperations requires CEOs to delegate significant au-thority to other managers. Kets de Vries and Miller(1984) found that CEO personality had a dramaticinfluence on SMEs, because CEOs frequently havedirect and personal contact with most levels ofmanagement. CEOs in SMEs play a vital role indetermining and reshaping strategy, dominate de-cision making, and set the climate of the firmthrough their style, goals, and attitudes. Severalempirical studies have also found strong relation-



9/25

ships between CEO personality, firm strategy, andperformance in SMEs (Miller & Droge, 1986; Miller& Toulouse, 1986). Moreover, young SMEs, such asthose in business process outsourcing (92 percentof sampled firms were less than ten years old), aretransitional, have short histories, and face few ofthe institutional and bureaucratic forces that dom-

inate larger and older firms. Therefore, CEO traitsexert a significantly stronger influence on firmstrategies than would be likely in the organization-al and institutional context likely to exist in olderand larger firms (Kets de Vries & Miller, 1984).Thus, we considered the SME nature of the Indianfirms to be well-suited to examination of our theo-rized relationships.

Fourth, our research design, which we con-structed to avoid common method bias and biasesresulting from using secondary sources to modelCEO personality (e.g., Peterson et al., 2003), re-quired us to have considerable access to CEOs,other top managers, and firms financial records.This setting also offered satisfactory access tothese difficult-to-obtain data.

Although use of samples from outside the UnitedStates is increasingly encouraged, internationalmanagement scholars are urging researchers to con-textualize their theoretical models deeply withinthe cultural context of the country studied (Tsui,2007). We discuss two specific facets of the Indiansociocultural context that are most relevant to un-derstanding the role of CEO personality in Indian

business process outsourcing firms: (1) the consis-

tency and relevance of the five-factor model inIndia and (2) the nature of the influence of CEOpersonality on firm strategies in the Indian socio-cultural context.

Consistency and relevance of the five-factormodel in India. Two central issues here arewhether the five-factor model has a similar mean-ing in India and in the U.S. and whether it shows apattern of relationships with expected traits and

behaviors in Indian samples that is similar to thatshown in U.S. samples. Several studies have foundscalar and factor structure equivalence in tests of

the model between Indian and U.S. samples (Judgeet al., 2002; Schmitt, Allik, McCrae, & Benet-Mar-tinez, 2007). These studies also demonstrated con-sistency in the magnitude of the five factors and intheir patterns of distribution across age and genderin Indian and U.S. samples. Finally, the five-factormodel predicted expected behaviors and traits (e.g.,self-esteem) in the Indian samples. These resultssuggest that the model not only has a similar mean-ing in India to its meaning in the United States butalso is central to understanding the relationships

between personality and behaviors in the Indian

context. Therefore, we expected our theoreticalpredictions based on empirical five-factor modelstudies conducted with U.S. samples to be valid inthe Indian context.

Influence of CEO personality on firm strategiesin the Indian sociocultural context. We expectedour theoretical predictions about the influence of

CEO personality on firm strategies to be both rele-vant and strong in the Indian sociocultural contextfor two reasons. First, although traditional studieshave labeled Indian culture as collectivist, withsmall cohesive social groups and emphases on fam-ily considerations and on collective rather thanindividual goals (Hofstede, 1980), recent studiesexamining cultural orientation at the individuallevel have found mixed results for India. For exam-ple, Sinha and Verma (1994) found that Indiangraduate students express more individualist orien-tations, with emphasis on independence, auton-omy, and individual goals, than collectivist orien-tations, as the result of Western influence,immediate life concerns, and exposure to mass me-dia. Similarly, Sinha, Sinha, Verma, and Sinha(2001) found that Indian students considered indi-vidual goals as important as or even more impor-tant than family and collective goals. Ghosh (2004)found that entrepreneurs and small business own-ers in India had significantly more highly individ-ualistic orientations than did other professionals,such as teachers. CEOs of Indian business processoutsourcing firms represent the educated (95 per-cent of sampled CEOs had completed at least an

undergraduate degree) and relatively young (with amean age of 37.12 years) entrepreneurs (90 percentof sampled CEOs founded their companies), whohave been shown to have a more individualisticorientation than the general Indian population.Therefore, we expected the CEOs of Indian SMEs toexercise autonomy and independence in strategicdecision making, with an emphasis on individualgoals rather than family and social considerations,making individual CEO characteristics, such aspersonality traits, more influential in strategic de-cision making than are broader social consider-

ations, such as family background and socioeco-nomic status.Second, as have earlier studies of Indian cul-

ture (Hofstede, 1980; Krishna, Sahay, & Walsham,2004), recent studies have shown that Indian busi-ness process outsourcing managers have a highpower distance orientation, implying an accep-tance of hierarchical authority and associated work

behaviors. For example, Levina and Vaast (2008)found that senior Indian business process outsourc-ing managers dominated decision making and in-teractions with clients, whereas lower-rank em-



10/25

ployees willingly accepted guidance and directionsfrom superiors and clients. Dibbern et al. (2008)also found high power distance behaviors such asdominance among information technology (IT) pro-fessionals in business process outsourcing firms. Ahigh power distance orientation of both CEOs andother employees is likely to increase CEO domi-

nance in strategy formulation and implementationin Indian business process outsourcing firms, com-pared with Western firms, in which CEOs are likelyto embrace decision-making styles reflecting an ori-entation to relatively low power distance. Thus,CEO personality is likely to exert a stronger influ-ence on firm strategies in Indian business processoutsourcing firms than in firms in the U.S. andother low power distance cultures.

Sample and Data Collection

First, we obtained a list of firms from the Feder-ation of Indian Micro and Small and Medium SizedEnterprises (FISME), which, with over 200,000members, is the largest SME association in India.We targeted SMEs in a large Indian city that hosts alarge number of business process outsourcingfirms. These firms varied in age (two to ten years),size (20 to 2,500 employees), type and range ofservices offered (e.g., customer interaction services,front- and back-office services), and clientele (e.g.,large and small businesses from Europe and NorthAmerica). Analysis of variance (ANOVA) based onFISME membership data revealed no significant

differences in demography between these firms andthose in other major cities in India. Thus, thesefirms represented a microcosm of the Indian off-shore business process outsourcing industry. Aftersample selection, we called each firm to ensure thatit was independent and that offshore business pro-cess outsourcing was its primary business, definedas at least 60 percent of sales coming from thissegment (Rumelt, 1974). We obtained a sample of427 independent firms with offshore business pro-cess outsourcing as their primary business.

We contacted the CEOs of the 427 firms by tele-

phone and asked if they and their top managerswould participate in the study by completing andreturning two questionnaires and by furnishingtheir recent financial performance data. The CEOsof 217 firms initially agreed to participate. We col-lected data at four different times through a desig-nated coordinator in each firm. First, we sent theCEOs a personality and demographic survey. Amonth later, after we had received the completedCEO surveys, we sent strategic flexibility surveys,to be filled out by at least two top managers report-ing directly to their firms CEO. All the scales were

in English. Finally, we requested financial perfor-mance records from each firms designated coordi-nator at two different time points: six months andone year following the receipt of the strategic flex-ibility surveys. We contacted the coordinator bytelephone to confirm the receipt of each question-naire as well as to remind him or her about the

return of the questionnaires. For 84 firms, we hadto conduct multiple follow-ups. We did not findany significant differences in the model and controlvariables between early- and late-responding firms.A total of 195 firms provided complete data, whichwe used in the analyses.

The 195 firms in the final sample did not differsignificantly from nonresponding firms (232) inage (F 1.41, n.s.), size (F 1.09, n.s.), owner-ship type (publicly held or privately held) (F 1.26, n.s), and range of service offerings (F 1.61, n.s.). We also used Heckmans (1979) two-step residual procedure to estimate selection biascaused by the nonresponding firms. The rho ( 0.11, s.e. 0.07, n.s.), sigma ( 0. 04, s.e. 0.02,n.s.), and Lambda/inverse Mills ratio (0.17, s.e.0.12, n.s.) were insignificant for the selectionequations. These statistical values suggest thatour sample was representative and did not sufferfrom nonresponse bias.

Measures

CEO personality. We measured personality viathe 60-item revised NEO Five-Factor Inventory (12

items for each factor) (Costa & McCrae, 1992), anextensively validated and used measure of the five-factor model (Costa & McCrae, 1988). Examples ofitems include I often feel inferior to others (emo-tional stability); I like to have a lot of peoplearound me (extraversion); I am pretty good aboutpacing myself so as to get things done on time(conscientiousness); I spend time reflecting onthings (openness to experience); and I am inter-ested in people (agreeableness). All items werescored on a scale ranging from 1 (strongly dis-agree) to 5 (strongly agree). We reverse-coded

the ratings on emotional stability to improve ourinterpretation of results. Coefficient alpha reliabili-ties were .79 for emotional stability, .70 for extra-version, .81 for conscientiousness, .74 for agree-ableness, and .72 for openness to experience.

Strategic flexibility.We measured strategic flex-ibility by adapting Grewal and Tansuhajs (2001)five-item scale assessing a firms ability to respondto environmental variations. Use of this scale wasconsistent with our conceptualization of strategicflexibility. The scale is conceptually robust, spe-cific to the strategic domain (unlike other scales



11/25

that address product development or technologyflexibility), and valid and reliable (Grewal & Tan-suhaj, 2001).

We pilot-tested the five strategic flexibility itemson 30 middle-level Indian business process out-sourcing managers who were not in our final sam-ple. After completing the pilot questionnaire, each

of these managers reviewed all questions for con-tent, clarity, meaningfulness, and construct mea-surement (Bagozzi, 1980). We also used item-totalcorrelation and discrimination based ont-statisticsto eliminate redundant items (Churchill, 1979).None of the five items had low item-total correla-tions. Given these results of the pilot test, we re-tained the five strategic flexibility items.

The five items of this scale were (1) We regu-larly share information and costs across businessactivities, (2) We frequently change our strategiesand structures to derive benefits from environmen-tal changes, (3) Our strategy emphasizes exploit-ing new opportunities arising from environmentalvariability, (4) Our strategy reflects a high level offlexibility in managing political, economic, and fi-nancial risks, and (5) Our strategy emphasizesversatility and empowerment in allocating humanresources. All items were scored on a scale rang-ing from 1 (strongly disagree) to 5 (stronglyagree). The coefficient alpha reliability for thescale was .84.

Firm performance. We used three establishedaccounting-based measures of firm performancefrom the financial records provided by the sampled

firms: return on assets (ROA), return on sales(ROS), and return on investment(ROI) (McDonaldet al., 2008). Because our sampled firms were pro-prietary or partnership firms, we could not includemarket-based measures of performance, such as theratio of book value to market value, or stock price.In the primary analyses, we lagged performance byone year after the survey date, but in separate anal-yses we used a half-year lag and found that thehypothesized results were unchanged.

Control variables. We used three firm demo-graphic characteristics (firm size, firm age,andfirm

past performance), three firm resource variables(R&D intensity, capital intensity, and advertisingintensity), three CEO demographic variables (CEOage, CEO position tenure, andCEO education), andTMT size as controls (Carpenter et al., 2001).Younger and smaller firms are more dynamic andtransient than older and larger firms, which tend to

become bureaucratic (Miller & Chen, 1996). Thus,younger and smaller firms are likely to shift theirstrategies frequently and then achieve greater flex-ibility than older and larger firms, which are likelyto focus on tried and true strategies and the status

quo (Miller & Chen, 1996). We measured firm age asthe number of years from a firms founding date to2005 (the year in which we collected the data). Wemeasured a firms size by the logarithm of the three-year average of its total number of employees(Guthrie & Olian, 1991).

Existing firm characteristics, such as capital in-

tensity, R&D intensity, and advertising intensity,represent important contingencies for developingfuture strategic flexibility. High capital intensity(capital expenses divided by sales) indicates afirms heavy investment in long-term assets, whichfosters strategic persistence rather than flexibility(Dess & Beard, 1984). High R&D intensity (R&Dexpenses divided by sales) and high advertisingintensity (advertising expenses divided by sales)imply heavy investment in innovation and productdifferentiation, which drive searches for new ideasand new ways of doing things in the future (Rajago-palan & Datta, 1996). Thus, capital intensity islikely to inhibit, whereas R&D and advertising in-tensity are likely to foster, strategic flexibility. Wemeasured these variables using data from the finan-cial records of the sampled firms for the year pre-ceding the date of collecting CEO personalityvariables.

Change in performance is an important determi-nant of strategic change (Greve, 1998; Rajagopalan& Spreitzer, 1997). An increase in performance re-inforces the value of existing strategies and resultsin maintenance of the status quo, whereas perfor-mance declines force managers to question the va-

lidity of existing strategies and foster changes instrategies. Thus, change in past performance islikely to relate negatively to strategic flexibility. Wemeasured the one-year change in past performanceusing ROA, ROS, and ROI (McDonald et al., 2008)for the year immediately preceding the survey date.

Greater CEO age has been associated with rigidityand resistance to change, whereas lower CEO agehas been associated with aggressive strategicchange (Wiersema & Bantel, 1992). Therefore,younger CEOs are likely to drive strategic flexibil-ity, whereas older CEOs are likely to inhibit it.

CEOs with long tenures develop set habits, estab-lish routine information sources, and rely largelyon past experience; high commitment to the statusquo and reluctance to consider strategic changeresult (Finkelstein & Hambrick, 1996; Wiersema &Bantel, 1992). Thus, CEOs with shorter positiontenures are likely to foster greater strategic flexibil-ity than CEOs with longer tenures. We measuredCEO position tenure as the number of years a CEOhad held the position at the time of data collection(Herrmann & Datta, 2002).

Previous studies have suggested that a high level



12/25

of education increases a CEOs receptivity tochange in corporate strategy (Wiersema & Bantel,1992); highly educated CEOs are likely to promotestrategic flexibility more than CEOs with relativelylower levels of education. CEO education level wasassessed by use of a seven-point scale (1 highschool, 2 attended college, 3 undergradu-

ate degree, 4

attended graduate school, 5

masters degree, 6 attended doctoral pro-gram, 7 doctorate) (Herrmann & Datta, 2002).

The greater size of a firms TMT, the greater thediversity of skills and perspectives it contains, andthis diversity is likely to stimulate strategic flexi-

bility (Eisenhardt & Schoonhoven, 1990). Follow-ing previous TMT research (Judge & Miller, 1991),we measured TMT size by asking each CEO toname the key managers who actively participatedin strategic decisions. This operationalization ofTMT size was based on the premise that the out-comes of a strategic decision are largely a functionof who participates in the decision-making process(Jackson, 1992).

ANALYSES AND RESULTS

We tested our theoretical model by use of struc-tural equation modeling (SEM; Joreskog & Sorbom,1993). We used LISREL 8 (Joreskog & Sorbom, 1993)to test our model in three steps. As recommended, weused mean-centered values of our construct measuresin the SEM analysis (Bollen, 1989). First, we usedconfirmatory factor analysis (CFA) to examine the

convergent validity of our construct measures asbased on the factor loadings of the individual mea-sures on their a priori defined factors.

Second, we examined the significance of thenonlinear relationship of agreeableness with strate-gic flexibility by comparing two structural models:1

a linear-only model and a nonlinear model. In thelinear-only model, we included only the linearagreeableness variables (without the squared agree-ableness term). In the nonlinear model, we in-cluded both the linear variables and a squaredagreeableness term (Bollen, 1989). We computedthe squared term from the mean-centered variables

of agreeableness.Third, we tested the mediation effects of strategicflexibility by comparing three alternative, nestedmodels: the fully mediated model (hypothesizedmodel), a partially mediated model (direct relation-ships of CEO personality variables to firm perfor-mance and indirect relationships of strategic flexi-

bility with the CEO personality variables and withfirm performance), and a nonmediated model (di-rect relationship of CEO personality variables tofirm performance). Following the recommenda-tions of others (e.g., MacCallum & Austin, 2000;Marsh, Balla, & McDonald, 1988), we used severalwidely accepted model fit adequacy indexes: thechi-square statistic, adjusted goodness-of-fit index(AGFI; Joreskog & Sorbom, 1993), incremental fitindex (IFI; Bollen, 1989), and root-mean-square er-ror of approximation (RMSEA). A significant im-provement in the fit of the fully mediated modelover the nonmediated and partially mediated mod-els would confirm the mediation effects of strategicflexibility. We report the results of these analysesin the following sections.

Validity and Reliability of Construct Measures

We validated the construct measures using CFA,which is most suitable for confirming whether con-struct measures load on their respective a prioridefined constructs (Browne & Cudek, 1993). Therange of loadings for the five personality factorswere as follows: conscientiousness, .81 to .92; ex-traversion, .78 to .94; agreeableness, .83 to .90; emo-tional stability, .75 to .90; and openness to experi-ence, .79 to .91. The factor loadings of the strategicflexibility measures ranged from .77 to .93, andthose of firm performance ranged from .84 to .95.

Table 1 shows descriptive statistics, correlations,and reliabilities for the eight construct measures.These results suggest high reliability and validityfor our study measures.

Nonlinear Model Fit Analyses

Our comparison of the linear-only (withoutagreeableness squared) and the nonlinear models(with agreeableness squared) indicated that thenonlinear model (2 97.14, df 30; AGFI .92,IFI .95, RMSEA .05) had a considerably better

1 For each model, we assigned one manifest variable(based on the participants average of the mean-centeredscale scores) to one latent variable for the five personalityfactors, strategic flexibility, and firm performance. As the

reliability estimates of manifest variables affect a modelsparameters, we fixed the error variances of the manifestvariables (Jreskog & Srbom, 1993). Error variance wascalculated via the reliability estimates (alpha coeffi-cients) presented in Table 1 (Joreskog & Sorbom, 1993:3738). This procedure allows an analysis of the struc-tural relations among the latent rather than the manifestvariables. However, we also tested our hypothesizedmodel using the three individual performance measures(ROA, ROS, ROI) rather than the averaged single measureof firm performance. These results were consistent withthe primary results (2 104.51; AGFI .90; IFI .92;RMSEA .05).



13/25

TABLE1

Descrip

tiveStatisticsandCorrelationsa

Variables

Mea

n

s.d.

Coefficient

Alpha

Reliabilities

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Controls

1.Firm

size

b

2.1

5

1.04

2.Firm

age

8.5

7

3.12

.24*

3.Pastperformance

0.0

8

0.10

.11

.05

4.Capitalintensity

0.7

8

0.21

.20

.22

.15

5.R&Dintensity

1.3

5

0.32

.14

.19

.20

.22

6.Advertisingintensity

1.1

4

0.29

.17

.12

.14

.18

.12

7.CEOage

32.1

2

5.42

.20

.13

.08

.17

.22

.16

8.CEOeducation

4.4

1

1.72

.09

.04

.20

.19

.20

.12

.09

9.CEOtenure

6.1

4

2.37

.19

.09

.07

.28**

.21

.17

.11

.12

10.TMTsize

3.0

1

2.16

.16

.10

.14

.11

.14

.11

.04

.07

.05

Modelvariables

11.Conscientiousness

3.4

8

0.74

.81

.07

.11

.19

.22

.20

.21

.10

.05

.09

.16

12.Emotionalstability

3.0

9

0.59

.79

.12

.08

.17

.12

.15

.11

.15

.09

.14

.07

.19

13.Agreeableness

3.6

9

0.47

.74

.05

.14

.15

.15

.15

.17

.11

.04

.10

.04

.15

.24*

14.Extraversion

3.7

5

0.52

.70

.17

.12

.21

.17

.20

.22

.07

.15

.18

.09

.09

.22*

.20

15.Opennesstoexperience

3.8

4

0.62

.72

.14

.09

.25*

.20

.24*

.20

.20

.08

.12

.05

.23*

.25*

.22

.23*

16.Strategicflexibility

3.3

4

0.74

.84

.18

.15

.26*

.25*

.23*

.20

.25*.14

.22*

.19

.22*

.37***

.20

.25*

.41***

17.Firm

performance

0.1

0

0.03

.85

.10

.16

.29**

.20

.24*

.22

.21

.17

.20

.21

.28**

.29**

.19

.30**

.32**

.44***

a

n

195.

b

Meanandmedianvaluesoffirm

sizearethenaturallogarithmictransfo

rmationsoftherawdata.Themeanfirm

numberofemployeeswas411.05and

themedianwas396.

p

.10

*p

.05

**p

.01

***p

.001


14/25

fit to the data than the linear-only model (2 139.71, df 33; AGFI .75, IFI .79, RMSEA .07). The chi-square difference between the twomodels was also significant (32.57, p .001, df 3), indicating that our hypothesized nonlinearmodel had a better fit with the data than the linear-

only model. We show the standardized structuralparameters of our hypothesized nonlinear model inFigure 1. Agreeableness squared relates negativelywith strategic flexibility ( 0.25, p .01),which confirms the inverted U-shaped relationship

between agreeableness squared and strategic flexi-bility and supports Hypothesis 3.

Mediation Model Fit Analyses

We tested our complete mediation model usingthe SEM approach suggested by James and Brett

(1984), which differs from the widely used incre-mental approach of Baron and Kenny2 (1986) intwo ways (for details, refer to James, Mulaik, andBrett [2006]). First, unlike the Baron and Kennyapproach, which uses a partial mediation model asthe base model, the SEM approach uses the more

parsimonious complete mediation model as itsbaseline. Thus, the SEM approach a priori excludesthe direct relationship between the independent

2 We tested our mediation model by means of theBaron and Kenny approach in a regression analysis usingcomposite average mean-centered measures for the per-sonality variables, strategic flexibility, and firm perfor-mance. These results were consistent with our primaryresults and supported our mediation hypotheses; theyare available from the authors upon request.

FIGURE 1Standardized Structural Coefficients for the Hypothesized Fully Mediated Modela

Conscientiousness

Emotional Stability

Agreeableness

Agreeableness Squared

Extraversion

Openness to

Experience

Strategic FlexibilityFirm

Performance

0.35***

0.29***

0.21*

0.25**

0.37

***

0.41***

0.34***

a The standardized structural coefficients for the control variables are as follows: firm size, 0.09; firm age, 0.05; past performancechange, 0.17*; capital intensity, 0.14*; R&D intensity, 0.20*; advertising intensity, 0.09; CEO age, 0.15; CEO education, 0.08; CEOtenure, 0.16; TMT size, 0.10.

*p .05**p .01

***p .001



15/25

variable and the dependent variable as a conditionfor mediation (James et al., 2006). Second, com-plete mediation is confirmed in the SEM approach

by explicitly testing the indirect relationship fromthe independent variable to the dependent variablethrough the mediator, rather than by testing thedecrease in the coefficient of the relationship be-tween the independent and the dependent variableonce the mediator is entered, as in the Baron andKenny approach (Shrout & Bolger, 2002). There-fore, in the SEM approach, the relationship fromthe independent variable to the dependent variableis not used as a control in estimating the relation-ship between the mediator and the dependent vari-able. Rather, mediation is indicated when the paths

between the independent variable (here, the CEOpersonality variables) and the mediator variable(strategic flexibility), as well as the path betweenthe mediator variable (strategic flexibility) and theoutcome variable (firm performance) are signifi-cant, and the overall model shows acceptable good-

ness of fit (James et al., 2006).The model fit indexes, which are presented in

Table 2, suggest an excellent fit for our hypothe-sized model (2 97.14,df 30; AGFI .92; IFI .95; RMSEA .05). The structural coefficients ofour hypothesized, fully mediated model (Figure 1)indicate that conscientiousness has a negative rela-tionship to strategic flexibility ( 0.35, p .001), whereas emotional stability relates positivelyto strategic flexibility ( 0.29, p .001). Bothextraversion ( 0.37, p .001) and openness tonew experience have a positive relationship to stra-

tegic flexibility (

0.41,p

.001). These resultssupport hypotheses Hypotheses 1, 2, 4, and 5. Asdiscussed earlier, the agreeableness-squared term isnegative and significant for strategic flexibility ( 0.25, p .01), indicating an inverted U-shapedrelationship and supporting Hypothesis 3. Strategicflexibility ( 0.34,p .001) relates positively tofirm performance. Thus, Hypothesis 6 is supported.Together, these results support the mediation con-ditions (James et al., 2006).

To further test the mediation hypothesis for stra-tegic flexibility, we compared our hypothesized

(fully mediated) model with the partially mediatedmodel and the nonmediated model, as recom-mended by Kelloway (1998). In the partially medi-ated model, we specified direct paths from the CEOpersonality variables to firm performance and in-cluded all other specifications in the basic hypoth-esized model. In the nonmediated model, we spec-ified direct paths from each CEO personalityvariable to firm performance and dropped the in-direct paths from the CEO personality variables tostrategic flexibility and from the strategic flexibilityvariables to firm performance.

The partially mediated model had a satisfactoryfit with the data (2 94.03, df 25, AGFI .80,IFI .82, RMSEA .07). However, the model-datafit is not as strong for the partially mediated modelas for the hypothesized model. The change in thevalue of chi-square between this model and thehypothesized model was also marginal and nonsig-nificant (2 3.11,df 1). Moreover, the addeddirect paths from conscientiousness ( 0.10,

n.s.), emotional stability ( 0.14, n.s.), agreeable-ness ( 0.09, n.s.), agreeableness squared ( 0.19, n.s.), extraversion ( 0.12, n.s.), and open-ness to experience ( 0.17, n.s.) to firm perfor-mance were not significant. The nonmediatedmodel did not fit the data well, with several in-dexes failing to meet the requirements (2 149.56,df 33, AGFI .69, IFI .64, RMSEA .11),results that are consistent with the a priori assump-tion of complete mediation in the SEM approach(James et al., 2006). These results indicate that thefully mediated hypothesized model had the best fit

and supported Hypothesis 7.We checked for interaction effects among the fivefactors. We did not find any interactions among thefive constructs in influencing strategic flexibility.

DISCUSSION

Our study yielded two major results: (1) eachvariable in the five-factor model of personalitymeasured for a firms CEO influenced the firmsstrategic flexibility and (2) strategic flexibility me-

TABLE 2Model Fit Indexes

Model FitNull

ModelNonmediated

ModelPartially Mediated

ModelFully Mediated

(Hypothesized) Model

2 (df) 371.23 (54) 149.56 (33) 94.03 (25) 97.14 (30)AGFI 0.50 0.69 0.80 0.92IFI 0.54 0.64 0.82 0.95

RMSEA 0.20 0.11 0.07 0.05



16/25

diated the relationship between CEO personalityand firm performance.

Theoretical Implications

Upper echelons and CEO psychology research.Our results contribute to the upper echelons and

CEO psychology theories in several ways. First,previous studies have not paid adequate atten-tion to the mechanisms underlying the relation-ships between CEO personality and firm perfor-mance (exceptions are studies by Gupta andGovindarajan [1984] and Peterson et al. [2003]).Our study extends upper echelons research bysuggesting howpersonality attributes of CEOs in-fluence firm performance. We specified a media-tor (strategic flexibility) through which CEO per-sonality influences firm performance. Our resultssuggest that the effectiveness of a CEO personal-ity trait depends on whether the trait enhances orinhibits strategic flexibility. CEO extraversion,emotional stability, and openness to experienceenhanced firm performance by fostering strategicflexibility, whereas CEO conscientiousness un-dermined firm performance by inhibiting flexi-

bility. Medium levels of agreeableness maxi-mized strategic flexibility and, consequently,firm performance. An important implication ofthis result is the need for empirical studies toidentify specific mediators in the relationship

between CEO personality and firm performance.Such studies are critical to developing a more

complete understanding of how CEO personalityattributes influence firm performance.

Second, our study demonstrates the importanceof the five-factor model of personality in a strategiccontext. Prior studies have examined attributes thatcapture only a narrow slice of CEO personality(e.g., locus of control) or that, despite intuitive ap-peal, lack strong psychological and methodologicalgrounding (e.g., CEO hubris) (Hiller & Hambrick,2005: 298). Researchers have been urged to usevalid frameworks from the psychology literaturethat comprehensively explain fundamental person-

ality differences in CEOs (Carpenter et al., 2004;Hiller & Hambrick, 2005). Cannella and Monroe(1997) noted that the CEO psychology literaturemay understated the contribution that personalitycan make to explaining the behavior of CEOs (andother top managers), because researchers havefailed to use comprehensive and robust frameworksof personality attributes in their studies.

The five-factor model, one such framework, pro-vides a valid, robust, and comprehensive way ofrepresenting fundamental personality differences(Judge et al., 2002). Despite this models rigor and

comprehensiveness, to our knowledge only oneprior study has used it to examine CEO personality(Peterson et al., 2003). Moreover, the authors of thatstudy examined the relationship of the personalityfactors measured for CEOs to TMT decision mak-ing, whereas we focus on strategic behavior (strate-gic flexibility). Our results highlight the relevance

of CEO personality factors measured to one strate-gic behavior and underscore the need for examin-ing their relationship to other strategic behaviors,such as innovation (Subramaniam & Youndt, 2005)and alliance formation (Eisenhardt & Martin, 2000).Such studies may strengthen the contributions ofCEOs personality attributes in explaining theirstrategic behaviors.

Third, studies examining specific CEO personal-ity attributes are sparse (Gupta & Govindarajan,1984; Miller, Kets de Vries, & Toulouse, 1982;Miller & Toulouse, 1986; Peterson et al., 2003).Most upper echelons studies have used demo-graphic variables as proxies for personality vari-ables (Finkelstein & Hambrick, 1996; Pitcher &Smith, 2001). We found weak correlations betweenCEO demographic characteristics (e.g., age, tenure,and education) and CEO personality attributes. Animportant implication of this result is that, forCEOs, demographic variables may not be appropri-ate proxies for personality variables. Our resultssupport recent criticisms of the use of CEO demo-graphic characteristics as proxies for CEO person-ality attributes (Carpenter et al., 2004; Lawrence,1997).

Personality research. Our results have impor-tant implications for personality research in thefield of organizational behavior. Our results foremotional stability, extraversion, and openness toexperience were consistent with published psy-chology and leadership research (Bono & Judge,2004; Judge et al., 2002). However, our resultsfor conscientiousness and agreeableness differedsomewhat from results of extant studies. Leader-ship studies have indicated that conscientiousnessand agreeableness relate positively to the effective-ness of team and functional leaders (Bono & Judge,

2004); our results indicate that conscientiousnessundermines firm performance by inhibiting strate-gic flexibility, whereas a medium level of agree-ableness maximized strategic flexibility and conse-quently firm performance.

The inconsistency in these results may have sev-eral explanations. First, it suggests that insightsabout psychological attributes in the psychologyand leadership literatures based on lower- and mid-dle-level managers may not always be replicatedfully in studies concerning CEOs. Our results sup-port the contention of the strategic choice (Child,



17/25

1972) and upper echelons (Finkelstein & Hambrick,1996; Hambrick & Mason, 1984) theories that deci-sion making at the top level (the strategic level) isunique and distinct from that at other levels in afirm. Thus, deemphasizing conscientiousness and

balancing assertiveness and altruism (mediumagreeableness) may be more critical for CEOs in-

volved in decision making at top levels than formanagers operating at other levels in the firm. Fu-ture theorists of strategic leader attributes maywant to consider the uniqueness of the strategiclevel.

Second, the negative impact of conscientious-ness on strategic flexibility and consequently onfirm performance could be due to the high mu-nificence (60 percent growth) (Tapper, 2004) inthe Indian business process outsourcing industryfor our study period that has occurred as a resultof the global outsourcing boom. This high munif-icence may have provided incumbent firms withthe confidence and energy to develop an aggres-sive, opportunistic, change orientation that isthen further validated by high performance(Lumpkin & Dess, 2001). Therefore, attributes ofconscientiousness such as dependability, perse-verance, and need for achievement served as bar-riers in this virtuous cycle of change propelled byhigh industry munificence. However, these pat-terns of relationships could be different underconditions of economic downturn eroding indus-try growth. In times of moderate or low munifi-cence, incumbent firms may shift to a more sta-

ble, cautious, and rational orientation (Van deVen & Poole, 1995) in which attributes of consci-entiousness such as dependability, achievement,perseverance, efficiency, and responsibility mayhelp the firms cope effectively with environmen-tal scarcity and improve rather than inhibit per-formance. Examining the relationships betweenconscientiousness, strategic flexibility, and firmperformance in periods of low munificence is animportant extension of our study.

Finally, the negative influence of conscientious-ness could have resulted from our focus on short-

term performance (we measured six-month andone-year lags). Recent literature on strategic flexi-bility suggests that although it generally has a pos-itive influence on performance in fast-changing en-vironments, the specific costs and benefitsassociated with strategic flexibility in the short andthe long run may differ (Johnson et al., 2003). Forexample, in the short run, an aggressively change-and flexibility-oriented strategy may yield superiorperformance. However, to achieve long-term suc-cess, firms need to balance tried-and-true strategiesand tighter control with change and risk taking

(Musteen, Datta, & Herrmann, 2009). This argumentsuggests that conscientiousness may be related neg-atively with short-term performance but may havean inverted U-shaped relationship with long-termperformance. Very high levels of conscientiousnessmay result in inertia and adverse performance,whereas very low levels of conscientiousness may

create instability and uncertainty for firms and, as aresult, firm performance may be maximized atmedium levels of conscientiousness. Examiningthe implications of strategic flexibility for long-term performance is an important area for futureresearch.

Managerial cognition. We based our hypothesesabout the relationships between the personalityfactors and strategic flexibility on the cognitivefiltering mechanisms described in research on

bounded rationality (Simon, 1991) and managerialcognition (Weick, 1995). The central contention inthis body of work is that firms are continuously

bombarded with complex and ambiguous informa-tion that is beyond the cognitive capacities of stra-tegic decision makers, who make sense of thisvastness and complexity by constructing mentalmodels as bases for strategic decision making. Un-like personality attributes, which are relatively sta-

ble, mental models are dynamic and changethrough learning (Cannella & Monroe, 1997; Ham-

brick & Fukutomi, 1991).An important area of research in bounded ra-

tionality is how strategic decision makers de-velop attention in mental models. Most scholars

in this area have theorized about the impacts ofindustry context (Nadkarni & Barr, 2008) andfirm context (Cho & Hambrick, 2006; Ocasio,1997) on the attention focus of managers. Forexample, Nadkarni and Barr (2008) found thatindustry velocity influenced whether the atten-tion focus of top managers was directed towardthe general or the task sector of the external en-vironment. Ocasio (1997) theorized that firmscommunication and procedural channels (e.g.,action memoranda, personnel evaluations, bud-getary and capital appropriations requests) affect

the attention of decision makers.Our results extend this literature by hintingthat personality attributes (measured per the five-factor model) of CEOs may influence their atten-tion focus, which in turn may influence strategicflexibility. Although we did not explicitly mea-sure mental models, the theoretical mechanismsthat we used to develop our hypotheses areclosely tied to attention, which is embedded inmental models (Bogner & Barr, 2000; Nadkarni &Narayanan, 2007). Thus, our results hint that at-tention may mediate the relationship between



18/25

personality factors and strategic flexibility. Thiscontention is consistent with cognitive psychol-ogy studies that have shown that although indi-vidual mental models are dynamic and changeover time through learning, relatively stable traitssuch as cognitive ability (Bieri, 1961; Meyer,1982; Ryan & Sackett, 1987) and emotional intel-

ligence (Goleman, 1995) influence them. For ex-ample, Goetzmann et al. (2007) found that five-factor model traits influenced the types andfrequency of metaphors salient in the domain-specific mental models of lung transplant pa-tients. Thus, these traits may influence the typesof mental models that CEOs develop, the fre-quency with which CEOs change their mentalmodels, and the patterns of changes in the mentalmodels. For example, CEOs with high opennessto experience may notice and absorb more newstimuli and thus may develop broad and complexmental models, as well as change their mentalmodels more frequently and more substantiallythan CEOs with low openness to experience. Ex-amining the relationship between personalityand attention in mental models is an importantarea of future research.

Strategic flexibility. Our results also contributeto the literature on strategic flexibility by highlight-ing the role of CEO personality in developing suchflexibility. This literature has focused on the influ-ence of technological (Evans, 1991; Sanchez, 1995;Worren et al., 2002), resource (Harrigan, 1980;Young-Ybarra & Wiersema, 1999), and network

(Young-Ybarra & Wiersema, 1999) structures onstrategic flexibility. Our results extend this litera-ture by suggesting that CEO personality is crucial.Our results are especially meaningful because weincluded several controls, including resource anddemographic variables that have been consideredas antecedents of strategic flexibility. An importantimplication of our results is the need for studies inthis area to focus on other CEO attributes that couldpotentially influence strategic flexibility. The coreself-evaluation (CSE) framework, for instance,identifies a significant and common core of four

attributes: self confidence, generalized self-effi-cacy, emotional stability, and locus of control(Judge, Thoresen, Pucik, & Welbourne, 1999). Re-cently, Hiller and Hambrick (2005) stressed the rel-evance of the core self-evaluation attributes to ex-plaining CEOs strategic behaviors. Future studiescould examine the relationship between these at-tributes and strategic flexibility.

Our results also have implications for the roleof industry context. Organizational behaviorstudies have suggested that relationships be-tween five-factor model traits and work outcomes

could be contingent on the dynamism of a taskcontext (Lepine & Van Dyne, 2001). The strategyliterature also suggests that effective adaptationto environment is different for firms in dynamicenvironments than it is for firms in stable indus-try contexts (Bogner & Barr, 2000; Eisenhardt &Martin, 2000; Nadkarni & Narayanan, 2007). To-

gether, these studies suggest that the relation-ships between five-factor model traits (especiallyconscientiousness and agreeableness), strategicflexibility, and firm performance could pan outdifferently in a stable industry context. Testingour model in a stable industry context is an im-portant area of future research.

Limitations and Future Directions

Our use of SMEs in this research limits the gener-alizability of the current results to large corporations.However, SMEs play a critical role in several high-technology industries, including electronics, aero-space manufacturing (Kaivanto & Stoneman, 2007),and biotechnology (Luukkonen, 2005). Moreover, re-sults of SME studies have made valuable contribu-tions to strategic theories such as those on knowl-edge-based resources (Wiklund & Shepherd, 2003)and internationalization (Oviatt & McDougall, 1995).Nonetheless, future studies should test these relation-ships in large corporatio

CEO Personality Nadkarni y Herrman (2010)

Documents