Hambrick 1996

The Influence of Top Management Team Heterogeneity on Firms' Competitive Moves
Author(s): Donald C. Hambrick, Theresa Seung Cho and Ming-Jer Chen

Source: Administrative Science Quarterly, Vol. 41, No. 4 (Dec., 1996), pp. 659-684
Published by: Sage Publications, Inc. on behalf of the Johnson Graduate School of Management,
Cornell University
Stable URL: http://www.jstor.org/stable/2393871 .
Accessed: 17/06/2014 08:50
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Sage Publications, Inc. and Johnson Graduate School of Management, Cornell University are collaborating
with JSTOR to digitize, preserve and extend access to Administrative Science Quarterly.
http://www.jstor.org
This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

All use subject to JSTOR Terms and Conditions
The Influence of Top This paper explores the executive origins of firms' com-
Management Team petitive moves by focusing on top management team
characteristics, specifically on team heterogeneity, rather
Heterogeneity on Firms' than on the more often studied environmental and orga-
CompetitiveMoves nizational determinants of such behaviors. Arguing that
competitive actions and responses represent different
Donald C. Hambrick, decision situations, we develop propositions about how
Theresa Seung Cho, heterogeneity may enhance some competitive behaviors
but impair others. With a large sample of actions and re-
and sponses of 32 U.S. airlines over eight years, we find re-
Ming-Jer Chen sults that largely conform to our propositions. The top
ColumbiaUniversity management teams that were diverse, in terms of func-
tional backgrounds, education, and company tenure, ex-
hibited a relatively great propensity for action, and both
their actions and responses were of substantial magni-
tude. Heterogeneous teams, by contrast, were slower in
their actions and responses and less likely than homoge-
neous teams to respond to competitors' initiatives. Thus,
although team heterogeneity is a double-edged sword,
its overall net effect on airline performance, in terms of
changes in market share and profits, was positive.*
In recent years, strategy researchers have directed attention
to the specific moves and countermoves made by rivals
within industries(e.g., Chen and MacMillan,1992; Smith,
Grimm,and Gannon,1992), recognizingthat organizational
vitalitymay depend more on the quantity,quality,and speed
of a firm's and its adversaries'on-going actions and re-
sponses in the competitivearena than on the traditionalcon-
cept of sustainableadvantage(D'Aveni,1994). Studies have
shown some importanteffects of firms' competitive behav-
iors on company performance,indicating,for instance, that
quick respondersgain marketshare at the expense of slow
responders (Chenand MacMillan,1992). Research also has
shed lighton the determinants,or antecedents, of firms'
competitiveaction and response tendencies. Forexample,
Chen and Hambrick(1995) demonstratedthat small and
largefirms differ markedlyin their competitive behaviors;for
example, small firms were shown to be more active and
faster in initiatingcompetitive challenges than large firms.
This paper, like Chen and Hambrick(1995), is one of a series
of studies drawnfrom an extensive and progressivelydevel-
oped database that represents a comprehensive longitudinal
study of domestic airlinecompetitionduringthe 1980s. In
contrastto previousstrategy research, which had tended to
infercompetitionfrom more static and aggregate firm attri-
butes or from structuralpropertiesof the industry,this pro-
gram of research has examined some fundamentalcompeti-
tive issues throughthe systematic analysis of concrete, tan-
C 1996 by CornellUniversity. gible market-orientedactions taken by firms. The research
0001-8392/96/4104-0659/$1.00.
programfalls into three distinct but highlyrelated streams.
0
The first stream, focusing on competitive dynamics or the
The authorsacknowledgesupportfrom
ColumbiaUniversity'sManagementInsti- action/responsedyad, has shown that competitive response
tute. WarrenBoeker,SaraKeck,John can be predictedby the attributesof the attack, the attacker,
Michel,Peter Murmann,MichaelTush- and the defender (Chenand MacMillan,1992; Chen, Smith,
man, and RuthWagemanmade helpful
suggestions on earlierdrafts.We are and Grimm,1992; Millerand Chen, 1994). The second
gratefulfor access to the airlinedata stream, of which the present study is a part, has investi-
base jointlyassembled by the thirdau- gated strategic competitive behaviorat the firm level-its
thor, MartinJ. Gannon,CurtisM. Grimm,
and KenG. Smith. humanand organizationaloriginsas well as performance
659/AdministrativeScience Quarterly,41({1996): 659-684

implications(Smithet al., 1991; Chen and Hambrick,1995).
The thirdstream, conceptualizinga firm's strategy as the en-
tire repertoireof its competitive moves, has examined the
antecedents and performanceimplicationsof competitive
inertia(Millerand Chen, 1994), simplicityof competitive
moves (Millerand Chen, 1996a), and nonconformityin com-
petitive repertoires(Millerand Chen, 1996b).
The cumulativegoal of this research programhas been to
develop a coherent predictivetheory of competitive behav-
ior,which has so far been lackingin the field of strategic
management. Such a theory should ultimatelybe more reli-
able if it is based on a common set of empiricalobserva-
tions. This research approach,akinto the PIMSstudies (e.g.,
Hambrick,MacMillan,and Day, 1982) and the Aston studies
(e.g., Pugh and Hickson,1972), is somewhat unusual in the
field of strategic management but should be an effective and
efficient way to accumulate knowledge in this field. The mul-
tiple studies-warranted because of the scale, detail, and
evidences of reliabilityof the data set-have been comple-
mentary,yieldingintegratedinsights into the complex topic
of competitive dynamics.
This stream of research is still embryonic,however, and far
more needs to be known. In particular,theorists of competi-
tive dynamics can benefit, we believe, by expandingtheir
attentionfrom environmentaland organizationaldetermi-
nants of behaviorto includethe characteristicsof the deci-
sion makers, in particularthe company's top management
team. Such an upper-echelonsapproach(Hambrickand Ma-
son, 1984) would acknowledge that humanand social bi-
ases, filters, and idiosyncraticprocesses at the top of the
organizationsubstantiallyinfluence competitive behaviors.
This line of thought has been consistently supported in ex-
aminationsof a wide arrayof organizationaloutcomes (sum-
marizedin Hambrick,1994), but with limitedexceptions
(Smithet al., 1991), it has not yet been considered as a way
to improveexplanationsof microcompetitiveaction and re-
sponse behaviors.
Takingthe upper-echelonsperspective, this paper examines
the effects of top management team heterogeneity on firms
competitive behaviorsand ensuing organizationalperfor-
mance. We expected that heterogeneity, a centralconstruct
in the literatureon top management, would be importantin
competitivedecision making,conferringbreadthof perspec-
tive, on one hand, but with the potentialfor team dissensus
and inefficiency,on the other hand (Jackson, 1992). With a
large sample of specific competitive actions and responses
of 32 U.S. airlinesover eight years, we examine the associa-
tions between team heterogeneityand (a) the firm's propen-
sity to act, the magnitudeof these actions, and the speed
with which they are executed, and (b) the firm's propensity
to respond to adversaries'actions, the magnitudes of re-
sponses, and the speed of responses.
BACKGROUND
Competitive Behaviors
Contraryto a model of sustainableadvantage, Schumpeter
and other theorists of the Austrianschool (Jacobson, 1992)
660/ASQ, December 1996

Team Heterogeneity
contended that competitionis fluidand that a firm advances

by continuallytakingaction and respondingto the actions of
its adversaries.One stream of strategy research has adopted
this frameworkby focusing on the microcompetitivebehav-
iors of firms and the determinantsand consequences of
those moves. Inthis view, firm performancecan be seen as
an outcome of a series of competitive actions that may be
rapidlynullifiedby opponents' incessant competitive chal-
lenges (D'Aveni,1994).
Because of the importanceof competitive behaviorsto over-
all firmstrategy and performance,an understandingof the
determinants,or influencers,of those behaviors is essential.
Several studies have alreadycontributedto this understand-
ing. MacMillan,McCaffery,and Van Wijk(1985) showed that
competitors'response times to new productswere a func-
tion of whether the productshad majororganizationalsignifi-
cance for these firms. And a series of studies, using some
of the same airlinedata we are using, has revealed addi-
tionaldeterminantsof competitive behaviors.First,using the
action/responsedyad as its analyticallevel, research has
demonstratedthat competitive response (likelihood,timing,
etc.) can be predictedby the attributesof the actions (Chen,
Smith, and Grimm,1992; Millerand Chen, 1994) as well as
those of the initiatorsand of the defenders (Chen and Mac-
Millan,1992). Second, distillingthe propertiesof a firm's ac-
tions and responses to captureits competitive behaviorat
the firm level, research has furthershown the significance of
such behavioralantecedents as firm size (Chen and Ham-
brick,1995), information-processing capacity(Smithet al.,
1991), and marketgrowth (Miller and Chen, 1994).
The research to date has focused almost solely on environ-
mental and organizationalexplanationsfor competitive be-
haviors.Except for Smith et al. (1991), whose study of the
impactof top management teams' educationaland industry
experience focused exclusivelyon competitive responses,
no attentionhas been paid to the decision makers behind
the observed marketbehaviors.Because research focusing
on top management team characteristicshas greatly aided
understandingof other organizationalphenomena (e.g., Ban-
tel and Jackson, 1989; Finkelsteinand Hambrick,1990), we
believe such a focus will also increase our understandingof
competitive behaviorsof firms, which represent instances of
specific decisions or bounded sets of decisions in rivalrous
arenas. Attentionto decision makers can providean impor-
tant new vantage for researchersof competitive interaction,
with possibly importanttheoreticaland practicalimplications.
Thus, with newly collected informationabout top executive
characteristicsadded to the database of competitive interac-
tions used in the priorstudies, the currentresearch delves
into the largelyunexploredterrainof the human origins of
competitive decision makingand behavior.
Althoughother characteristicsof top management teams
might also fruitfullybe examined, we restrictour theoretical
focus here to heterogeneity,first, to elaborateon the com-
plex effects of this importantconstruct and, second, be-
cause the fundamentaldifferences between a firm's com-
petitive actions and its responses to competitors' actions
providea uniquearena for studyingthe potentiallydivergent

effects of heterogeneityand possibly understandingconflict-
ing results in the literatureon top management teams.
Heterogeneity of Top Management Teams
Theoreticalinterest in top management teams can be traced
to Marchand Simon's behavioraltheory of the firm (March
and Simon, 1958; Cyertand March,1963). Accordingto this
theory, decision makers are often unable to make economi-
cally rationaldecisions because they are boundedlyrational
and must act in a social context of multipleand conflicting
goals. Hambrickand Mason (1984) extended these ideas in
their upper-echelonsperspective. In their view, (1) the orga-
nizationbecomes a reflectionof its top executives, and (2)
the characteristicsand functioningof the top management
team have far greater potentialfor predictingorganizational
outcomes than do the characteristicsof the chief executive
officer (CEO).
Numerousstudies over the last ten years have found signifi-
cant associations between the demographiccomposition of
the top management team and organizationalcharacteristics.
Studies have documented the tendency for young, short-
tenure, highlyeducated teams to be relativelyinnovative,
even after controllingfor the type of industry(Banteland
Jackson, 1989; O'Reillyand Flatt,1989). Organizationalten-
ure of top management team members was found to be
stronglyassociated with strategic persistence, or absence of
change (Finkelsteinand Hambrick,1990; Grimmand Smith,
1991; Wiersema and Bantel, 1992). Other studies have
found organizationaleffects arisingfrom the mix of func-
tionalbackgrounds(e.g., Thomas, Litschert,and Ra-
maswamy, 1991), industryexperience (Eisenhardtand
Schoonhoven, 1990), and turnoverof top management
teams (Virany,Tushman,and Romanelli,1992). Moreover,
top management team characteristicsconsistently predict
organizationaloutcomes better than do CEOs'characteristics
alone (summarizedin Hambrick,1994). Scholars of top
teams have become particularlyinterested in the effects of
the team's heterogeneity, the variationin team members'
characteristics,which has been called a theoreticalfulcrum
for research on groups and top management teams (Jack-
son, 1992).
Althoughmany investigationshave been conducted on the
effects of heterogeneity in groups in general (reviewed by
Jackson, 1992) and several on top management teams in
particular(reviewed by Hambrick,1994), the conclusions
have been contradictory.Beneficialeffects have been ob-
served in some studies. Banteland Jackson (1989) found
that top management team heterogeneity in educational
level and functionalbackgroundwas positivelyassociated
with innovativenessin a large sample of banks. Eisenhardt
and Schoonhoven (1990) showed that growth rates of semi-
conductorcompanies were positivelyrelated to the top
team's heterogeneity in industrytenure. The imputed logic
for these positive effects has been the same as originally
proposed by Hoffmanand Maier(1961): diversityenhances
the breadthof perspective, cognitive resources, and overall
problem-solvingcapacityof the group.
6621ASQ,December 1996

Team Heterogeneity
Other studies have found negative effects from top team

heterogeneity. O'Reillyand Flatt(1989) showed that com-
pany innovationwas negativelyrelatedto team heterogene-
ity in firmtenure in a wide cross section of firms. O'Reilly,
Snyder,and Boothe (1993) similarlyfound that team hetero-
geneity in firmtenure was negativelyrelated to adaptive
change in a sample of electronics firms. Althoughheteroge-
neity may providewider cognitive resources, it may also cre-
ate gulfs or schisms that make the exchange of information
difficult(Anconaand Caldwell,1992). In some instances, het-
erogeneity may engender outrightdistrustand acrimony,as
widely dissimilargroup members may have differentvocabu-
laries, paradigms,and even objectives. Thus their aggregate
cognitive endowment can become a net liability,with organi-
zationalinnovationand performancesuffering. Furthersup-
portingthis latterlogic, Wagner, Pfeffer, and O'Reilly(1984)
found that heterogeneity in firmtenure was associated with
higherturnoverin top management teams, and the more
distant an individualexecutive was from the average of the
other group members, the more likelyhe or she was to de-
part.Jackson et al. (1991) obtainedthe same result in a
sample of banks, furthersuggesting that demographicdiver-
sity may lessen the social integrationof the top team. Mea-
suringtop team processes more directly,O'Reilly,Snyder,
and Boothe (1993) found that heterogeneity in tenure was
negativelyassociated with a multi-itemmeasure of team rap-
port. Similarly,Smith et al. (1994) found that heterogeneity in
tenure was negativelyassociated with informalcommunica-
tion within the team but had no association with social inte-
grationor communicationfrequency.Theirstudy thus illus-
trates the potentiallycomplex implicationsof top team
heterogeneity.
Directpsychologicalmeasures of heterogeneity, although
often used in research on work groups (e.g., Hoffmanand
Maier,1961), are very difficultto obtainfor senior executives
in majorfirms and are unavailablefor past, disbanded top
management teams. Thus, in line with almost all priorre-
search on top management team heterogeneity, we rely on
demographicconceptions of the group, accordingto which
the executives' functionalbackgrounds,educationalexperi-
ences, and firmtenures serve as proxies for their perspec-
tives, beliefs systems, and networks and affiliations(Jack-
son, 1992; Hambrick,1994).
While distinctionsmight be drawn between differenttypes
of demographicheterogeneity, no established theory for do-
ing so presents itself. Jackson (1992) noted the need to dis-
tinguishbetween differenttypes of heterogeneity, drawinga
distinctionbetween heterogeneity of "personalattributes"
(e.g., race, gender, and personality)and "task-relatedattri-
butes" (the specific skills and abilitiesneeded to performthe
job).This is a promisingline of thought but may have limited
applicationfor most research on top management teams,
since the majordimensions for describingteam members
reflect a combinationof skill-basedand personal attributes.
For instance, an executive's primaryfunctionalbackground
reflects his or her domain of professionalexpertise, but it
may also greatlyreflect the person's personality,cognitive
style, values, and other factors. The same could be said for
educationalbackgroundand tenure.

We expect each of the three forms of demographichetero-
geneity we examine to contributeindependentlyto disper-
sion in the group's perspective and that each will tap an
amalgamof differences in professionalorientations,values,
cognitive styles, business and life experiences, and networks
(Jackson, 1992). Althoughthe three dimensions are common
in priorresearch on top team heterogeneity, they have not
heretofore been examined simultaneously.
Heterogeneity and Competitive Behaviors
A firm's competitive behaviorsconsist of its moves and
countermoves, or its actions and responses to adversaries'
actions. Each move represents an instance of a decision, or
a bounded set of decisions, in which multiplesenior execu-
tives are typicallyinvolved.This is not to say that every com-
petitive move involves all members of the top management
group.A decision about a geographicexpansion may involve
all the firm's senior executives, whereas a decision about a
new sales force structuremay involveonly a subset of the
top team. But even in decisions involvingonly a partof the
firm,a numberof executives would likelyhave inputto the
decision, especially in single-business firms like airlines,in
which the activitiesof the senior-most executives are highly
interdependent.Thus, even though some of the top groups
we studied may not have had all the qualities needed to be
called "teams," as discussed by Hambrick(1994), their
members can be expected to interact,negotiate, and influ-
ence each other extensively, a conception that is in line with
priorupper-echelonsresearch.
The top team can thus be considered as the aggregate infor-
mationaland decisional entity throughwhich competitive
moves are made. These moves depend on the team's scan-
ning of the environment;recognizingproblems and opportu-
nities and interpretingother externalstimuli;developing po-
tential moves; negotiating,refining,and selecting moves;
and implementingdecisions (Mintzberg,Raisinghani,and
Th6oret, 1976; Fredricksonand laquinto,1989). Because top
management team heterogeneity can be expected to influ-
ence these internalprocesses in ways that substantiallyaf-
fect competitive behaviors,we examine the relationships
between top management team heterogeneity and three
majordimensions of a firm's actions and responses: (1) its
competitive propensity,or its relativetendency to undertake
actions and respond to competitors'actions, (2) the competi-
tive magnitude,or the relativescale and significance of the
firm's actions and responses, and (3) competitive speed, or
how quicklythe firm implements its actions and responds to
others' actions. In doing this, we distinguishbetween ac-
tions and responses as qualitativelydifferenttypes of deci-
sion situationsthat are affected in differentways by top
management team heterogeneity.
Competitive actions as creative initiatives. To undertakea
competitive action, not directlypromptedby an adversary's
priormove, a firm must initiateand invent to "create" the
move (MacMillan,1982). Some actions are unequivocally
novel, such as AmericanAirlines'initialfrequent-flyerpro-
gram. Othercompetitive actions, such as expandinginto
new geographicmarkets,are less innovativebut are still acts

Team Heterogeneity
of initiative.In the context of competitive interplay,actions-

comparedwith responses-can be considered creative initia-
tives. It is in creative, unstructuredtasks that top manage-
ment team heterogeneity is expected to have its most
positive effects (Jackson, 1992). A diverse team has broader
cognitive resources, encompassing a wider field of vision
and more extensive externalcontacts, than does a homoge-
neous team (Hoffmanand Maier,1961; Jackson, 1992).
With multifacetedbackgroundsand orientations,heteroge-
neous top management team members can observe more
opportunities,threats, and overallstimulion multiplefronts
and thus have a broaderpotentialrepertoirefor generating
actions.
At the same time, because of its diversity,the heteroge-
neous team may experience internalconflict and strains,
which could result in slow decisions. It might furtherbe ex-
pected that the conflictthat often accompanies diversity
would impairthe firm's abilityto launchactions at all. Over-
all, however, the heterogeneous team has access to such
wide-rangingstimuliand has such a broadpotential reper-
toire that its abilityto conceive and launchactions on many
fronts should outweigh the dampeningeffects of internal
strains. Undertakingcompetitive actions is foremost a func-
tion of being able to create, or generate, those actions, par-
ticularlyin very turbulentindustrieslike the airlineindustry
that lack clear models of competitive behavior.The top man-
agement team's abilityto agree on certainactions, while not
unimportant,is a secondary determinantof the firm's action
propensity.Therefore,comparedwith homogeneous teams,
which have relativelyrestricted,redundantscanning and ac-
tion-designcapabilities,the heterogeneous team can be ex-
pected to engage in quantitativelymore actions than homo-
geneous teams. Thus, we propose:
Proposition 1: Top managementteam heterogeneity is positively
relatedto the firm'saction propensity.
The magnitude,or significance,of an action is also impor-
tant. Some moves are relativelysubstantialin scale and bold-
ness, while others are incrementaland routine(MacMillan,
McCafferey,and VanWijk,1985). The magnitudeof an ac-
tion can be characterizedby its strategic significance, its
noteworthiness, and its scope. The strategic significance of
a firm's actions can be gauged by the firm's relativeuse of
strategic ratherthan tacticalinitiatives.Strategic moves in-
volve large outlays, long time-horizons,great departures
from the status quo, and are difficultto reverse (Ghemawat,
1991; Chen, Smith, and Grimm,1992). Tacticalmoves, such
as price changes and sales force redeployments, tend to be
incrementaland limitedin their implications.Competitiveac-
tions similarlyvary in their noteworthiness, or the amount of
attentionthey receive from industryobservers and analysts
(Porter,1980; Chen and Hambrick,1995). Some moves re-
ceive a great deal of attention in an industry,often because
they are radical,creative, or complex, while others receive
bare mention, often because they are conventional,limited,
and predictable.Finally,a firm's actions can be considered in
terms of their scope, or the extent of the firm's operations
that are affected by the moves (Chenand MacMillan,1992).
Initiativestaken only in one productline or one region are

relativelynarrowin scope, while others affect the company's
full range of operations,all of its productsand markets (Por-
ter, 1980). We expect that these three elements of competi-
tive magnitude-strategic significance, noteworthiness, and
scope-will reflect the firm's degree of boldness and, indi-
rectly, innovationin its competitive initiatives.The firm that
undertakesstrategic and highlynoteworthy moves of broad
scope is behavingdifferentlyfrom a firm that takes primarily
tactical, little-noted,and narrowlyrestrictedmoves.
We expect top management team heterogeneity to contrib-
ute substantiallyto these differences. As a source of broad
cognitive resources, team heterogeneity has long been
thought to enhance creativityand innovation(Hoffmanand
Maier,1961; Banteland Jackson, 1989; Jackson, 1992).
Marketinginnovations,such as AmericanAirlines'introduc-
tion of a frequent-flyerprogramin the early 1980s, have
been attributedto top management team diversity(Labich,
1990). Chagantiand Sambharya(1987) similarlyobserved an
association between the functionaldiversityof top manage-
ment teams in tobacco companies and strategic adaptation.
With a broadarrayof informationand experience, diverse
teams can generate a wider range of options that synergisti-
callycombine the members' orientationswhile avoiding
groupthinkand behavioralinertia(Janis, 1972; Hambrick,
1994). Thus, we propose the following:
Proposition 2a: Top managementteam heterogeneity is positively
relatedto the strategic significanceof the firm's actions.
Proposition 2b: Top managementteam heterogeneity is positively
relatedto the noteworthinessof the firm's actions.
Proposition 2c: Top managementteam heterogeneity is positively
relatedto the scope of the firm'sactions.
The speed of organizationalmoves has also received increas-
ing attention (Smithand Grimm,1991; D'Aveni,1994), be-
cause the firm's speed in makingits moves allows it to
achieve earlyadvantages in its initiatives,often puttingits
competitionon the defensive (Stalk,1988; Eisenhardt,
1989). Althoughwe expect top management team heteroge-
neity to have positive effects on action propensityand mag-
nitude,we expect that it will slow an organization'sspeed in
executing its actions. Communicationsand decision making
in a heterogeneous team are cumbersome because of the
disparateperspectives and vocabularies(O'Reilly,Caldwell,
and Barnett,1989). Heterogeneitycan even be a source of
outrightinformationblockage and conflict (Zengerand
Lawrence, 1989; O'Reilly,Snyder,and Boothe, 1993). At a
minimum,the wide-ranginginformationand opinions avail-
able in a heterogeneous team requiretime to process (Jack-
son, 1992). In contrast,the homogeneous top management
team may take fewer actions, but when it does-when po-
tentialactions fall within the team's field of vision, prefer-
ences, and repertoire-it can act very quickly.We particularly
expect that top management team heterogeneity will affect
the amount of time taken to implement an action:
Proposition 3: Top managementteam heterogeneity is negatively
relatedto the firm'saction execution speed.
Overall,then, we expect the diverse, abundantcognitive re-
sources of heterogeneous top management teams to be

Team Heterogeneity
manifested in a propensityto launchcompetitive actions

with relativelygreat magnitude,but not quickly.Conversely,
homogeneous top management teams will undertakefewer,
less substantial,but faster actions than heterogeneous
teams. It is also possible that heterogeneitywould increase
the amount of time requiredto formulatean action, but we
do not have data to test this idea.
Competitive responses as induced counterplay. Compared
with launchinginitialactions, respondingto an adversary's
act does not requireas much abilityto scan broadlyand cre-
ate wide-rangingalternatives.When the adversaryhas taken
an action, providinga distinctand usuallyobvious stimulus,
the primaryquestion becomes, "Shouldwe respond?" For
two reasons, we expect management team heterogeneity to
have negligibleeffects in determiningthe firm's tendency to
respond. First,the stimuliare relativelyclear. Drawingon
Mischel's (1968) concept of "situationalstrength," when the
situationis "weak"-complex, ambiguous, and open to
many interpretations-the characteristicsof the actor greatly
affect his or her behavior,just as top management team het-
erogeneity is expected to affect takingaction. But when the
situationis "strong"-unambiguous, with few and clear-cut
stimuli-as in respondingto a competitor,then the stimuli
prevailin determiningbehavior.Even though there are in-
stances of secretive competitive actions (Prescott and
Smith, 1987), in an industrywith an active business press
and strong trade associations, competitors'actions can be
assumed to be universallyknown to all top teams. Because
the stimuliare generallyunambiguousand overt, then, a
firm's tendency to respond will be more a function of its
resources and other characteristics(Smithet al., 1991) than
of the cognitive limitsand biases of the top management
team.
Second, top management team heterogeneity will have little
effect on a firm's propensityto respond because there is
little need to create. The adversary'saction provides a tem-
plate, minimizingthe degree to which respondingrequires
idiosyncraticdesign capabilities.In its action, the initiatorhas
alreadyhanded the potentialrespondera possible response.
Even a homogeneous top management team, with limited
cognitive breadthand repertoire,can, at a minimum,often
simply imitatethe action. Thus, we propose:
Proposition 4: Top managementteam heterogeneity is unrelated
to the firm's response propensity.
We see a qualitativedifference between the decision to re-
spond and the characteristicsof the response itself. Al-
though situationalstrength may prevailin a firm's decision to
respond, once the decision has been made, we expect top
management team characteristicsto affect the design and
execution of the particularresponse. Heterogeneous top
management teams should draw on their broadercognitive
resources in designing more significant,bolder responses
than homogeneous teams, who could be expected to under-
take more incremental,measured, and conventionalre-
sponses-perhaps simply imitatingthe action-reflecting the
group's narrowand redundantcognitive endowment. Thus,
we expect heterogeneous top management teams to en-
gage in relativelyhigh-magnituderesponses. In assessing

response magnitude,noteworthiness and scope are the im-
portantindicators.Significance,or the proportionof re-
sponses that are strategic ratherthan tactical, is not useful
in the way it is for actions, because strategic actions tend to
engender strategic responses, and tacticalactions tend to
engender tacticalresponses. Thus, we propose:
Proposition 5a: Top managementteam heterogeneity is positively
relatedto the noteworthinessof the firm's responses.
Proposition 5b: Top managementteam heterogeneity is positively
relatedto the scope of the firm's responses.
As with actions, responses by a heterogeneous team are
likelyto be slower than those of a homogeneous top man-
agement team. The design, negotiation,and elaborationof
high-magnituderesponses from heterogeneous teams take
time. Added to the time such teams lose as a result of diver-
sity and consideringdifferentpoints of view is the time that
accrues strictlybecause the team often opts to design a
special response from its wide-rangingcognitive resources
ratherthan to counterwith a narrow,marginal,or imitative
response. In this study, we focus on two dimensions of re-
sponse speed: response generationspeed, or the time
taken to formulateand announce a response to a competi-
tor's action; and response execution speed, or the amount
of time taken to implementan announced response. We
propose:
Proposition 6a: Top managementteam heterogeneity is negatively
relatedto the firm's response generationspeed.
Proposition 6b: Top managementteam heterogeneity is negatively
relatedto the firm's response execution speed.
Thus, we expect heterogeneityto exert the same effects on
responses taken as on actions, increasingtheir magnitude
but slowing their development and execution. In contrast to
priorstudies that have sought primarilyto determine the
overalladvantages or disadvantagesof top management
team heterogeneity,we argue that heterogeneity may bene-
fit some particularbehavioraloutcomes but impairothers.
The next question is whether the benefits outweigh the
costs.
Heterogeneity and OrganizationalPerformance
Researchers have found inconsistent evidence about the ef-
fects of top management team heterogeneity on organiza-
tionalperformance:positive effects (Eisenhardtand
Schoonhoven, 1990), negative effects (Murray,1989), and no
effects (Micheland Hambrick,1992). It is clear from our own
propositionsthat heterogeneity may be a double-edged
sword. On the one hand,we expect heterogeneity to en-
hance a firm's action propensityand its action and response
magnitudes, potentiallybenefitingcompany performance
(Chenand MacMillan,1992). On the other hand, we expect
heterogeneityto reduce the firm's speed, both in acting and
responding,possibly leadingto lower performance(Eisen-
hardt,1989; Chen and MacMillan,1992).
Some theorists have incorporatedcontingency factors as po-
tential moderatorsof the heterogeneity-performancelink.
Jackson (1992), followingSteiner (1972), proposed that het-
erogeneity has benefits for unstructured,novel tasks, but
.6681ASQ,December 1996

Team Heterogeneity
homogeneity is better for routinetasks. Anotherdistinction,

drawn by Hambrickand Mason (1984), was between stable
and turbulentenvironments.In their view, heterogeneity
yields benefits in turbulentenvironments; homogeneity is
beneficialin stable environments.The U.S. airlineindustryin
the periodimmediatelyfollowingderegulationexperienced
exactly the conditionsunderwhich these theorists expect
top management team heterogeneityto be advantageous.
There was a high level of uncertaintyand turbulence,no
clear role models to be imitated,and extremely rivalrous
conditions(Levine,1987). Breadthof perspective, creativity,
and a willingness to undertakeunprecedentedaction were
requiredunderthese conditions,and a heterogeneous top
management should have been able to meet those needs.
Thus, we propose:
Proposition 7: Top managementteam heterogeneity is positively
relatedto overallperformanceimprovementsin a turbulentindus-
try.
RESEARCH METHODS
Sample
Datawere gathered from AviationDaily,the majorindustry
publication,on competitive moves by 32 majorairlines,
those noted by the Departmentof Transportationas having
annualoperatingrevenues of $100 millionor more, for the
post-deregulationyears of 1979 to 1986.
AviationDaily,a 50-year-oldindustryjournal,offered com-
plete and detailed informationon airlinecompetition.The
journalaims at objectivelyreportingairlines'announcements
and actions, which minimizesany concerns over post hoc
rationalizationof competitive moves and bias toward cover-
ing only certainairlines'activities.An extensive survey of 57
senior airlineexecutives and industryexperts (e.g., consult-
ants and analysts) indicatedthat the respondents considered
AviationDailycomprehensive,accurate, and a significant
source of informationfor the airlinesthemselves.
The method used in collectingthe data was similarto Miller
and Friesen's (1977), which has been labeled "structured
content analysis"(Jauch,Osborn,and Martin,1980). The
methodology is unique in that competitive interactionsof
sample firms were directlyidentifiedfrom an extensive re-
view of publicinformation.A specially designed, structured
coding schedule was used to performthe content analysis.
Dependent Variables
A competitive move has the potentialeffect of acquiringri-
vals' marketshares or reducingtheir anticipatedreturns.To
identifysuch moves, an extensive review of every issue of
AviationDailywas undertakento discover all of the competi-
tive moves in this industry,such as promotionalactivities or
marketexpansions, as suggested by Levine (1987). To iden-
tify the actions that met with responses, a researcherthen
identifiedall entries in AviationDailythat were responses by
searching for the following key words: "in respondingto,"
"following,""match,""underpressure of," or other words
that indicateda move was a response to a competitor's ac-

tion. The identificationof responses was straightforwardand
involvedlittlejudgment.
Next, three researcherscarefullytraced streams of actions
and responses back to the initialaction. First,the research-
ers read all AviationDailyissues in chronologicalorderto
find all competitive moves. Second, using the above key-
words, the coders first identifiedresponses and then worked
back to find the reportof the initialaction, thereby tracing
every initialaction and all of the announced responses to it.
In total, 1,445 moves (1,027 actions and 418 responses)
over the eight years were identified.They were classified
into 21 types (e.g., price cut, promotion,route entry),to al-
low statisticalcontrols for types of moves taken.
Actionpropensitywas calculatedas the total numberof ac-
tions a firm launchedin a given year. Response propensity
was measured by the total numberof responses a firm exe-
cuted in a year, given that it was identifiedas a competitor
affected by an action. To define the response opportunities,
the researchersfirst identifiedall airportsaffected by the ini-
tial action and defined the airlinesthat competed via those
airportsas affected and, hence, as potentialresponders.
They then calculatedresponse propensityas the percentage
of response opportunitiesthat the firmactuallytook in each
year.
Five measures were used for the competitive magnitude of
actions and responses: strategic significance, noteworthi-
ness, and scope of a firm's actions; and noteworthiness and
scope of responses.
Action significancewas measured as the numberof strate-
gic actions the firmtook in a year, dividedby its total num-
ber of actions. Strategicactions involve substantialinvest-
ments in fixed assets, people or structure,while tactical
ones involve smaller investments, without large-scaleor irre-
versible commitment (Millerand Chen, 1994). Three raters
workingindependentlyclassified actions as strategic (e.g.,
mergers and acquisitions,hub creations)or tactical (e.g.,
price cuts). There was total agreement among the raters in
classifyingthe varioustypes of moves as strategic or tacti-
cal.
Action noteworthiness was measured by the amount of in-
dustryattentionaccorded a move. We first counted the
numberof lines AviationDailydevoted to reportingthe ac-
tion, but, because press attentiondiffers inherentlyby type
of move, we standardizedthe noteworthiness scores for
each of the 21 types over all years and all airlinesto have a
mean of zero and standarddeviationof one. Similarly,we
measured response noteworthiness by the numberof lines
AviationDailydevoted to the response, standardizedfor
each type of response. We assume that AviationDaily,a
longstanding,well-respected industrypublication,allocates
space in proportionto newsworthiness and that, on average,
items commandingmore attentionare viewed as relatively
innovative,bold, unexpected, or complex.
Action scope, or the extent to which the action affects the
full breadthof the firm's operations,was measured as the
proportionof the firm's revenue base potentiallyaffected by

Team Heterogeneity
the action. After identifyingthe airportsaffected by a firm's

action, we dividedthe firm's annualnumberof passengers
at these airportsby the firm's total numberof passengers in
that year. Forexample, an airline'sdecision to initiatea na-
tionwide price discount would be greater in scope, affecting
all of its domestic market,than a discount at a specific air-
port. Response scope was similarlymeasured as the per-
centage of the firm's revenue base affected by its response.
Forcompetitivespeed, we used three measures: action exe-
cution speed, response generationspeed, and response exe-
cution speed. Action execution speed is the amount of time
a firmtook to execute an announced action. First,we mea-
sured the time between the date the firm publiclyan-
nounced or acknowledgedthe intended action and the date
that action began to be executed, as both were reportedin
AviationDaily.We then controlledfor the differenttypes of
actions by standardizingthe time lag across all years and
firms (mean of zero and a standarddeviationof one). To rep-
resent speed ratherthan time lag, the signs of the scores
were reversed so that a large value indicates great speed.
Response generationspeed was measured as the amount of
time a firmtook to generate a response if it did so. Relying
on the reportsin AviationDaily,we calculatedthe time lag
between the day that a competitorannounced its initialac-
tion and the day that the focal firmannounced its response.
This was again standardizedby response type and reverse-
scored to indicatespeed. Response execution speed was
the amount of time that a firmtook to execute an an-
nounced response. Because a firm's response execution
speed depends on the type of the initialaction and other
features that may have affected the execution time of that
initialaction, we regressed response execution time on type
of action and action execution time for all years and airlines
to controlfor these effects. We then used the average resid-
ual scores of the above regression results across all the re-
sponses made by an airlinein a given year as its response
execution time for that year. The signs of scores were again
.reversed so that a large value implies great speed.
To explore effects on organizationalperformance,we used
two measures. The first was the firm's marketshare change
duringthe year, calculatedas the percentage change from
the previousyear in marketshare in those markets in which
the airlinewas present (this year's marketshare minus last
year's, dividedby last year's). The second was the firm's
change in profitsduringthe year, calculatedas the percent-
age change in dollarprofitsbetween this year and the previ-
ous year.
Independent and Control Variables
Eachtop management team was defined as all executives
above the vice-presidentlevel (e.g., senior vice president,
vice chairman,chief executive officer).This operationaliza-
tion, used in priortop management team studies (e.g.,
Chagantiand Sambharya,1987; Murray,1989; Michel and
Hambrick,1992), yields a more complete group than if only
executives who are inside directorsare included(e.g., Finkel-
stein and Hambrick,1990) but a more restrictivegroup than
if all officers are included(e.g., Wagner, Pfeffer, and

O'Reilly,1984). The chosen operationalizationis intended to
includeall majorline and staff executives, encompassing all
the key activities of the firmat the senior-most level. This
definitionyielded a mean top management team size of 8.6
(standarddeviationof 3.4) executives. All informationpertain-
ing to management characteristicswas obtained from each
year's Dun and BradstreetDirectoryof CorporateManage-
ment.
We examined three types of top management team hetero-
geneity: functionalheterogeneity, educationalcurriculumhet-
erogeneity, and company tenure heterogeneity. Each of
these three types of heterogeneity has been used in prior
studies, but typicallynot together. The three types are com-
plementary,reflectingdiversityon somewhat differentdi-
mensions. To the extent that results are consistent across
the three types, we can have greater confidence in the ef-
fects of heterogeneity on competitive behavior.Functional
backgroundsrepresent executives' primaryprofessional ori-
entations, includingtheir implicitcausal models, vocabular-
ies, and internaland external networks (e.g., Hambrickand
Mason, 1984; Walsh, 1988). Educationalcurriculumsimilarly
reflects professionalorientationsbut also taps some funda-
mental formativequalitiessuch as cognitive style, values,
and early upbringing(Smartand Pascarella,1986). Company
tenure, which is stronglycorrelatedwith industrytenure and
age in our sample, indicates an executive's cohort, frame of
reference for viewing the historyof the firm, and internal
network (Wagner,Pfeffer, and O'Reilly,1984). We did not
measure tenure on the team, because we viewed it as a
less fundamentalreflectionof an executive's orientationthan
tenure in the company. Since top management team bound-
aries are permeable, membershipon the top management
team, as we defined it, may or may not have been an impor-
tant factor in an executive's point of view and internalsocial
network,at least not when comparedwith the importanceof
membershipin the firm.
Functionalheterogeneity was measured by a variationof the
Herfindal-Hirschman index,
16
H= 1- ,p2
where H is the heterogeneity measure and p the percentage

of top management team members in each of 16 functional
backgroundcategories listed in the Appendix.In almost all
cases, coding executives' functionalbackgroundswas
straightforward.In those few cases in which an executive
had substantialexperiences in more than one category, we
selected as the primarycategory the one in which the exec-
utive spent the most time. In a validitytest, Barbosa(1985)
found that 82 percent of executives indicatedtheir primary
functionas the same one he had coded from the Dun and
BradstreetDirectory.
Foreducationalbackgroundheterogeneity, we used eight
differentdisciplines,listed in the Appendix,to code each ex-
ecutive's educationalbackground.Forexecutives with gradu-
ate degrees, we coded the correspondinggraduatedisci-
plines; for those without graduatedegrees, we coded their

Team Heterogeneity
undergraduatedisciplines.We then calculatedthe Herfindal-

Hirschmanindex for educationalbackgroundheterogeneity.
Companytenure heterogeneity was calculatedas the stan-
darddeviationof the numberof years the executives on the
top management team had spent in the firm. We did not
use the coefficient of variation,since one of our controlvari-
ables was the top management team's mean company ten-
ure.
We also includedas controlvariablestop management team
size, average educationallevel (mean numberof years of
highereducation),and firm size (measured by annual reve-
nue-passenger miles). Firmsize is a particularlyimportant
controlvariable,because a numberof our dependent vari-
ables are inherentlya functionof company scale (Chen and
Hambrick,1995). Also, because competitive behaviors may
depend on resource availability(Smithet al., 1991), we in-
cluded three measures of organizationalslack: equity/long-
term debt, currentassets/current liabilities,and priorprofits
(alllagged by one year).These data were collected from
COMPUSTAT tapes. Finally,the years 1980-86 were in-
cluded as dummy variables,with 1979 as the omitted cate-
gory. The controlvariablesare those identifiedas most es-
sential in the competitive interactionand top management
team literatures.
Analysis
In analyzingthe data, we pooled the observations cross sec-
tionallyfor all years. Althoughthis approachproduces more
precise estimates, two potentialproblems exist: First,slope
coefficients may be inconsistent over time, makingthe
pooled techniques inappropriate.Second, the ordinaryleast
squares (OLS)estimates may be biased, since neglected
firm-specificcharacteristicswould lead to violatingassump-
tions about the independence of observations (Pindyckand
Rubinfeld,1991). Because a numberof tests, includingthe
Durbin-Watsontest and Bartlett'stest, indicatedserial corre-
lationand heteroskedasticity,we used a generalized-least-
squares technique for analysis (Harrigan,1982; Smith et al.,
1991).
We used Kmenta's(1986) autoregressive heteroskedastic
model to correctfor serial correlationand heteroskedasticity.
To correctfor first-orderserial correlationwe appliedthe
Cochrane-Orcutt transformationto each individualfirm and
correctedfor heteroskedasticityby dividingdependent and
independentvariablesby the firm-specificerrorvarianceob-
tained from the regressions on the serial-correlation-cor-
rected data. After these transformations,the data were
pooled and analyzedusing OLS regression tech'niques.This
approachwas used in a numberof recent related studies
such as Millerand Chen (1994), Smith et al. (1991), and
Finkelsteinand Hambrick(1990).
RESULTS
Descriptivestattisics and correlationcoefficients for all vari-
ables are presented in Table 1.

O ~~~CN
l~~~~~~~~~~~~
00
C4
N 0
o,
r- J
O.o.Oo
M N4 O CY) N CY)a) ( CY
)
o CN O. O. O.
C LO 0lt mCD
o - CD 0 00 O CD 0 o
cs S S bt LO oo c4 o 0O
O 00 OD - C14
dr- 'It Od - C1 0' m a) a) r- t L
CO O NCDO 0 O O O 7 CNooOC
I I I I I I I I I I
r%. '' C1) - 000LO- -00 0 0 0 C N
LO 00 Y)
C0 0000-000N0 0 0 0 OO O
C C? C(?) C c C 0)OC1(0 C L( 1 0) )
aB) a) I'* 00 CN CY)O i 00 CD 'It C-4 C-4 C-4 LO 'Iolt - O - M 01

OOOCO O OOO O OOO-CN O00000000
0 0
O
~ ~ ~ CY
~ ~ 0~
C O(t) CNC0 (t)
0~CNLCBOOO 0 C. O OOCN
OO O~ -NO (N LO0C
OCN
O)1) r M CD 0 0
C'l CN ~~~~~'t 04
CM
CY) O
0B(Clq
O O O O
Y
O O O O
CDdqdBt
O O O O
CD
()cO()
O C14 O O O O O
0~~~~~~~
1S 3rt
s tD- *---**' 13
m wm 00 M 00 CY) LO LO CY) 1 > d CN l CY) LO CD r- CB1N
0 0) CY
10 C IS00) CD 0I 00 00 aa1 000 o-o 0 C)
0 C00D
7 r- ) m mv m
CD 0 0 dm O mv ) ) -7 C -7 )O)
C? N
LO C)C) r
00 ~ ~ ~ ~ ~ ~ ~ ~ ~ 0 ~ N ON
U)
X I 0
-a - 00)h- =
D:t
-
Ct )C O8 )CD ? () { llo -0) ) 0v CD0( . i a
ln E 2
?~~~CN~C CY) m
LO~~~~-
C~~~~)(')
O C 00000 0 c 0o0 00 n?OB
CW 0-CON ?o?O-0 M '-h-'-0 ?O 0rorE 0 - 00 CO
(V) Q00D-C-
07-00 (N 10
I~( 0 0 000 OON0OON 0 C
9-9
O I I I0 I O
0) 0~~~~~1 U | .
o 0~~~~~~~~~~~~04 0-C
CD-) ) C C 00C- COC 0 0 -
(5 0)~~~~4
0)~~~~~~~~~~-6 )Z - )c -u
)
a 0 a
C 0a ) ) c L ) a) () a
NC N C-4> CY) Q)C ~~~~~~~f 0 O

U)
-o n (
U- .-.)
f) U n
0 f U) CN N
nUo)0 ) I )
(U .~~~~~~~~~~~~~. q
0 ~~~~~~~~~~~~
~~~~~~~~~~~~~~-'
0-oc U)
.) :
C.)
0 . ua a
0
~~~~~~~~~~~~~~~~~~~~~~~~~~~;-
ao )
(
a
;z Y
0D
'z U
-()
) 0 ;-;
aa)
.-
0
;
a ) a a0
U)
U)
co c-2)
m m t
>a) >c= -a ) a C- a)C-) -) ) O a)
2-U)Ua))~ a) a) ?-U()O U) 8U) -U) a) U) 0)
0c 0 c c0 -0 D
CO
<E< < m< <
coC4CO 4. U 6 - o ) C - -Lo( , <6a c1 1
c-
< Cr<<C
0 )O o )o
r<C
r-)0U)
)Mu
c)
D(
c-4
- (U--- C)0)4)H<< W>Z-<0-<-)<-E-C-OC- --- -l- <C < C < E C C CE4Cfl0 V1

Team Heterogeneity
Competitive Behaviors
Table2 presents GLSregression results, with action charac-
teristics as the dependent variables.Turningto the top man-
agement team heterogeneityvariables,the results for action
propensitygenerallysupportedproposition1. Both education
heterogeneityand tenure heterogeneity were significantly
positivelyrelatedto action propensity.
Table 2
Regression Results of Top Management Team Characteristicsand Control Variables on
Action Characteristics*
Action Characteristics
Execution
Variable Propensity Significance Noteworthiness Scope speed
Functionalheterogeneity .122 .063--- .114-- .076" - .010---

(.482) (.024) (.054) (.037) (.003)
Educationalbackground
heterogeneity .355--- - .053 .1 33- .022" - 144--
(.051) (.192) (.060) (.011) (.055)
Tenureheterogeneity .084-- .012 .234 -.237 -.122
(.040) (.051) (1.347) (.942) (.582)
Firm size .832---- .231---- - .133--- -.132 -.042
(.124) (.021) (.030) (.562) (.159)
Slack (E/LTD) -.053 -.012 -.424 -.073 .124
(.623) (.248) (1.542) (.442) (.562)
Slack (CA/CL) -.275-- .044--- .093 -.044 .221-
(.132) (.017) (.454) (.257) (.133)
Slack(Profit) .054---- .022---- - .123- .084 .023
(.015) (.005) (.051) (.281) (.245)
TMTsize -.022 -. 123-- -.132-- -.326- -.122
(.162) (.058) (.052) (.181) (.468)
Average education .022--- .134- - .424 - .314--- -.143
(.008) (.066) (.242) (.116) (.649)
Averagetenure -.321-- .064 .643-- .233 .334
(.152) (.242) (.298) (2.154) (1.382)
N 156 153 642 642 603

F 124 164 142 180 165
*p < .10; *-p < .05; Up < .01; ....p < .001.
* Because generalized-least-squares regressionequationscorrectedfor serialcorrelationare used, R2statistics are not
reported.Standardizedregressioncoefficientsare listed,with standarderrorsin parentheses. Because no systematic
patternexisted, the statistics for the dummyvariablesfor the years 1980-86 are not shown.
Propositions2a-2c linkingtop management team heteroge-

neity with action magnitudereceived substantialsupport.
Functionalheterogeneitywas significantlypositively related
to action significance,partiallysupportingproposition2a.
Proposition2b was more stronglysupported,with both func-
tional heterogeneity and educationalheterogeneity significantly
positivelyassociated with action noteworthiness. Proposition
2c also received general support;the signs for both func-
tional heterogeneity and educationalheterogeneity were
both positive and significant.Thus, the more diverse the top
management team, the greater the overallmagnitudeof the
firm's competitive actions.
Providinggeneral supportfor proposition3, both functional
heterogeneity and educationalheterogeneity were signifi-
cantly negativelyrelatedto action execution speed. Tenure
heterogeneity, however, did not have a significanteffect.

The controlvariablesshowed highlysignificantassociations
with the action characteristics,as they did in all the results
we will present. To maintainour focus, we will not comment
here on the controlvariablesbut will note the most signifi-
cant results below.
Table3 presents GLSresults for the response characteris-
tics. In contrastto proposition4, which predictedno relation-
ship, all three types of top management team heterogeneity
were significantlynegativelyassociated with response pro-
pensity. Thus, the more heterogeneous the top management
team, the less likelythe firmwas to respond to a competi-
tor's action.
Table3
Regression Results of Top Management Team Characteristicsand Control Variables on
Response Characteristics*
Response Characteristics
Generation Execution
Variable Propensity Noteworthiness Scope speed speed
Functional heterogeneity - .015--- .184-- .032-- - .542--- - .322--
(.006) (.089) (.015) (.180) (.159)
heterogeneity - .054-- .154 .125 - .245- - .025--
(.027) (.649) (.632) (.146) (.012)
Tenure heterogeneity - .155-- - .045 .342- - .386-- - .132
(.079) (.371) (.182) (.192) (.449)
Firm size .664---- -.123 -.320-- .465000 -.524----
(.154) (.554) (.150) (.156) (.075)
Slack (E/LTD) - .214---- .132 .234 .3620 - .213
(.041) (.848) (.944) (.211) (.792)
Slack (CA/CL) -.012---- -.032 -.222 .186 .211
(.002) (.250) (1.325) (.577) (1.078)
Slack (Profit) - .063--- .244 .086 - .285-- - .233-
(.025) (.911) (.641) (.143) (.138)
TMTsize -.126--- -.034 -.081 .225 -.053
(.042) (.152) (.365) (.762) (.792)
Average education .204-- - .122 - .323 - .153 .053
(.081) (.494) (1.597) (.706) (.782)
Average tenure .144---- -.243 -.224 53400 .124
(.032) (1.066) (.885) (.265) (.712)
N 139 286 286 263 263
F 723 484 155 401 112
*p < .10; *p < .05; ..e p < .01; .... p < .001.
* Because generalized-least-squares regressionequationscorrectedfor serialcorrelationare used, R2statistics are not
reported.Standardizedregressioncoefficientsare listed,with standarderrorsin parentheses. Because no systematic
patternexisted, the statistics for the dummyvariablesfor the years 1980-86 are not shown.
Propositions5a and 5b, on response magnitude,were par-

tiallysupported. Functionalheterogeneitywas significantly
positivelyrelatedto response noteworthiness and response
scope. Tenure heterogeneitywas also positivelyassociated
with response scope, but with marginalsignificance (p <
.10). Overall,though, the results tend to support the premise
that heterogeneous top management teams engage in rela-
tively significant,bold competitive responses.
Propositions6a and 6b posited that heterogeneous top man-
agement teams would be slower than homogeneous teams
in their competitive responses. Results show consistent sup-
6761ASQ,December 1996

Team Heterogeneity
port for these ideas. Allthree types of heterogeneity were

significantlyand negativelyassociated with response genera-
tion speed. Similarly,both functionalheterogeneity and edu-
cationalheterogeneitywere significantlynegatively related to
response execution speed. In the aggregate, there is evi-
dence that the heterogeneous teams were slower than ho-
mogeneous teams in their competitive responses.
Performance
We proposed that in the turbulentindustrywe are studying
that the advantages of top management team heterogeneity
would outweigh the disadvantagesand that the effects on
overallcorporateperformancewould be positive. The re-
sults, shown in Table 4, strikinglyconform to these expecta-
tions. The dependent variablesare the two performance
measures, growth in marketshare and growth in profits.
Performancefor the prioryear is includedas a controlvari-
able, to assure that the analysis is capturingstrictlychange
in performancein the focal year.
Table4
Regression Results of Top Management Team Characteristics and
Control Variables on Performance*
Growth in Growth in
Variable market share profits
Functionalheterogeneity .332---- .623.
(.042) (.298)
heterogeneity .064-- .3230
(.029) (.150)
Tenureheterogeneity .093- .3230
(.058) (.095)
Firm size .232 -.312.
(.594) (.142)
Slack(E/LTD) .144 .332
(.561) (1.389)
Slack(CA/CL) .561-- .084
(.224) (.334)
Slack(Profit) .524- .124
(.280) (.541)
Previousyear's performance
(growth in market share or profit) -.052-- -.134
(.025) (.492)
TMT size .353 -.212
(1.272) (.721)
Averageeducation .230-- .454-
(.115) (.252)
Averagetenure .083 .304
(.342) (1.492)
N 139 138
F 54 43
p < .10; *-p < .05; *--p < .01; *---p < .001.
* Because generalized-least-squares regressionequationscorrectedfor serial
correlationare used, R2statistics are not reported.Standardizedregression
coefficientsare listed,with standarderrorsin parentheses. Because no sys-
tematic patternexisted, the statistics for the dummyvariablesfor the years
1980-86 are not shown.
The results reveal a clear pattern:All three types of hetero-

geneity were positivelyassociated with performanceim-
provement, both in marketshare and profitability.Thus, de-

spite the low response propensityand slowness of the
heterogeneous top management team, its other benefits ap-
pear to more than compensate, and in general the airlines
with diverse top management teams advanced in their com-
petitive arena.These results show that top management
team heterogeneity is of considerableconsequence not only
to the firm's competitive behaviorsbut also to its perfor-
mance.
Effects from ControlVariables
A brief review of the effects of the controlvariablesis war-
ranted,particularly since they showed some strong associa-
tions with both competitive behaviorsand performance(as
seen in Tables 2, 3, and 4). Because the four variablesrepre-
senting firm size and slack have been includedin priorstud-
ies using the airlinedata, we will not comment on them. In-
stead, we will note brieflythe effects of the top
management team controlvariables.
Top management team size was negatively related to re-
sponse propensityand to three measures of competitive
magnitude(actionsignificance,action noteworthiness, and
action scope). This consistent patternmay suggest that large
teams tend to be restrainedin their competitive initiatives.
Team size was positivelyrelatedto growth in marketshare,
however, indicatingthat, overall,there may be some bene-
fits from large teams. Team members' average education
level was positivelyrelatedto action propensity,action sig-
nificance,and response propensitybut negatively related to
action noteworthiness and action scope. The average educa-
tion level of the top management team was positively re-
lated to both performancemeasures, however, indicating
overallbenefits from top management team education lev-
els. Finally,the average tenure of the team showed limited
effects, with little pattern.Average tenure was negatively
relatedto action propensity,but positivelyrelatedto action
noteworthiness, response propensity,and response genera-
tion speed. These results indicatethat the top management
team characteristicswe controlledfor may have substantial
implicationsfor competitive behaviorsand firm performance
and may thus meritfurtherstudy.
DISCUSSION
Both longstandingtheory (Schumpeter,1950) and more re-
cent empiricalevidence (e.g., Chen and MacMillan,1992)
indicatethat firm performancedepends greatly on the on-
going competitive behaviorsof the firm and its adversaries.
It is thus importantto improveunderstandingof the determi-
nants of those behaviors.This paper complements previous
work emphasizingthe environmentaland organizationalde-
terminantsof firms' microcompetitivebehaviorby incorporat-
ing the compositionof the top management team as an im-
portantinfluence in such outcomes. Seeking to explore in
depth the effects of one centrallyimportantupper-echelon
construct,top management team heterogeneity, we found
broadand significantassociations with both the firm's com-
petitive behaviorsand performance.
*678/ASQ,December 1996

Team Heterogeneity
Top Management Team Heterogeneity and

Competitive Behavior
The effects of top management team heterogeneity on the
firm's competitive actions conformedto our propositions.
First,heterogeneitywas positivelyrelatedto action propen-
sity, or the basic tendency to undertakecompetitive initia-
tives. Second, it was positivelyrelatedto the magnitudeof
competitive actions. Across multiplemeasures, gauging the
significance,noteworthiness, and scale of actions, we found
consistent evidence that heterogeneous top management
teams were bolderin competitive actions than homoge-
neous teams. Finally,we found the heterogeneous teams
were slower in their action execution than homogeneous
teams. These findingsfullyalignwith priortheory positing
several consequences of group heterogeneity: broadgather-
ing of information,decision creativityand boldness, but fric-
tion and slowness in decision makingand action.
Results for the firm's response characteristicswere partlyin
line with expectations but diverged in one importantway.
Whereas we proposed that there would be no association
between top management team heterogeneity and response
propensity,we found a negative relationshipfor all three
measures of heterogeneity.Thus there was broadlyconsis-
tent evidence that homogeneous teams were most likelyto
respond to their adversaries'initiatives.Our logic in propos-
ing no relationshipwas based on the premise that competi-
tors' actions are usuallyso clear-cutand unambiguousthat
they amount to "strongsituations"(Mischel, 1977) for the
potentialresponder,and top management team characteris-
tics will not influencethe likelihoodof a response. The find-
ing of a negative relationshipbetween heterogeneity and
response propensitycauses us to conclude that we were
perhaps consideringonly the scanning element of the re-
sponse decision-makingprocess and ignoringother elements
of the process (Mintzberg,Raisinghani,and Theoret, 1976).
It may be that initiators'actions are generallyunambiguous
in their occurrence,and homogeneous and heterogeneous
top management teams are equallyaware of such events,
but actions may be open to varyinginterpretations(Porter,
1980). Perhaps,then, heterogeneous teams experience dis-
sensus about issues rangingfrom the meaning of the com-
petitor'saction to the design of an appropriateresponse.
Conversely,the homogeneous team may have an advantage
in this respect. Its internalsimilarity,shared vocabulary,and
relativelyfluidexchange propertiesenhance its abilityto in-
terpretthe competitor'smove and decide to make a coun-
termove.
When acting, a top management team must develop a move
from its repertoire;as our results suggest, the broaderthe
repertoire,the greater the numberof actions that can be cre-
ated. In responding,however, the team becomes anchored
to the natureof the initialaction and does not have to create
anythingto respond. Simple imitationis commonly a possi-
bility.The negative association between heterogeneity and
response propensitysuggests that the diverse orientationsin
heterogeneous teams may lead them to engage in more in-
terpretation,negotiation,and creative formulationof re-

sponses, which lessens their likelihoodof respondingat all,
comparedwith the more straightforward,reflexive response
behaviorsof homogeneous teams. Such a portrayalis further
supportedby our other findingson response characteristics.
When heterogeneous top management teams did respond,
their responses were of greater magnitudebut slower than
those of homogeneous teams. These findingswere consis-
tent with our propositionsand add to the overallindication
that heterogeneous top management teams are cognitively
rich.
It might be said that heterogeneous teams are good (but
slow) at creating,but that homogeneous teams are good at
deciding (Jackson, 1992). Self-initiatedactions involve creat-
ing competitive moves, and top management team hetero-
geneity will aid in this. Respondingto another's actions,
however, involves littlecreating; it is primarilya matter of
deciding,and top management team homogeneity may en-
hance this competitivetendency. In this sense, our findings
are consonant with Eisenhardtand Schoonhoven's (1990)
mooregeneral conclusionthat heterogeneity of industryexpe-
rience broughthelpfuldiversityto the top teams they stud-
ied, but that long tenures helped group processes. In our
study, average team tenure, which is generallyseen as a
contributorto fluidgroup processes, had a very strong posi-
tive effect on response propensity,supportingthe notion
that top management groups with shared experiences and
perspectives are most able to react to clear-cutstimuli.An
additionalreason that top management team heterogeneity
dampens response propensitymay be that diverse teams
draw on their cognitive breadthto develop highlyelaborate,
multidimensionaloffensive strategies, which in turn restrict
their flexibilityin respondingto others' initiatives.Moreover,
such top management teams might become highlycommit-
ted to their elaborate,creative offensive strategies and un-
willingto move into a reactive mode.
Heterogeneity and OrganizationalPerformance
Ourresults indicatethat top management team heterogene-
ity may have potentialbenefits and drawbacksto the firm's
competitive behaviors.It is understandable,then, that prior
research has producedmixed results on the influence of top
management team heterogeneityon overallorganizational
performance.In this study of the airlineindustry,the advan-
tages of heterogeneityexceeded the disadvantages.All
three types of heterogeneityexamined were independently
and positivelyassociated with performance,measured both
as change in marketshare and change in profits.And each
type of heterogeneity contributedin its own way to overall
airlineperformance.
We would not assert that top management team heteroge-
neity is always beneficial.The U.S. airlineindustryin the pe-
riodstudied was exceedingly turbulentand lacked clear com-
petitive role models, precisely the conditions underwhich
theorists expect top management team heterogeneity to be
advantageous (Hambrickand Mason, 1984; Jackson, 1992).
It may be that a more stable industry,with more widely ac-
cepted models or recipes for behavior(Spender, 1989) and
less intense rivalrywould have favoredtop management
68O/ASQ,December 1996

Team Heterogeneity
team homogeneity. It is also possible that after the period

we studied, the airlineindustrybecame more orderlyand
programmed,yieldingdiminishedbenefits of top manage-
ment team heterogeneity. But our results indicatethat in a
highlyturbulentsetting, diverse top management teams, de-
spite their drawbacks,have superiorcompetitive and adap-
tive capabilities.
It is also importantto note that the three types of heteroge-
neity examined showed considerableconsistency in their
effects on competitive behaviorsand performance.Among
the 12 models presented in Tables 2, 3, and 4, two or three
of the heterogeneity measures were significantin the same
directionfor 10 of the models. There were no instances of
opposing significanteffects. Moreover,for those criterion
constructs for which we have multiplemeasures (such as
competitive magnitude,speed, and performance),the ef-
fects of heterogeneitywere highlyconsistent in their signs
across those indicators.Thus, even though we have exam-
ined three differentdimensions of heterogeneity-so differ-
ent that they are not correlatedamong themselves and
hence do not warrantbeing combined into an index-they
exhibitstrikinglyconsistent effects on corporateoutcomes.
These three forms of top management team dispersion may
differ in their subtleties and the specific conduits by which
they affect outcomes, but at the broadest level they can be
considered complementaryand corroborativeof the impor-
tance of top management team heterogeneity as a robust
construct.
Because our data were archival,we were unable to gain an
in-depthunderstandingof the actual processes of competi-
tive decision makingin the firms we studied. For instance, it
is possible that some of the actions and responses we stud-
ied were not a result of interactionsamong the entire execu-
tive group but, rather,involveda small subset of those man-
agers. Additionally,some decisions, such as tactical pricing
responses, may have been formulatedin a routine,even au-
tomated way that minimallyinvolvedthe senior executives.
The strong and consistent results we obtained, however,
suggest that these conditionsdid not prevailand that the
compositionof the top management team is of substantial
importanceto the competitive functioningof the firm. In-
sights about the actualteam dynamics and processes in-
volved in competitive moves will requirefield methods such
as those used by Eisenhardtand Bourgeois (1988), Gersick
(1989), and Smith et al. (1994). Our results suggest that field
work, directed at understandingthe actual processes of
competitive decision makingengaged in by homogeneous
and heterogeneous top management teams, could bear fruit.
Ourfindings illuminatethe importanceof focusing on execu-
tive characteristicsand processes as majorinfluences on
competitive dynamics.This conclusion is perhaps not so sur-
prising,consideringthat these microcompetitiveactions and
responses, often the buildingblocks of sustainable advan-
tage, tend to occur in a limitedtime span and with imperfect
information.The cognitive and social processes and biases
of the decision makersthemselves will almost invariablybe-
come manifested in their strategic choices in such situa-
tions.

In additionto highlightingthe "humanfactor" in interfirm
competitive behavior,our study may have importantpractical
implications.Forinstance, our results suggest that compa-
nies whose main focus is to launchinnovativeactions in an
effort to gain competitiveadvantage may wish to ensure
that their top management teams comprise executives with
diverse backgrounds.Conversely,firms that strive to be
promptlyreactiveto others' attacks may want to configure
their top teams with greater homogeneity so that maximum
efficiency in decision-makingresponsiveness and speed can
be achieved. Dependingon the environmentalcontext and
the firm's overarchingstrategic repertoire,the profileof the
top management team can be modifiedto make it more
congruentwith the firm's objectives. Overall,our findings
indicatethe need to incorporatethe composition of the top
decision makers in attempts to understandand enhance
competitive advantage.
REFERENCES
Ancona, Deborah G., and David F. Chen, Ming-Jer, Ken G. Smith, Fredrickson, James W., and An-
Caldwell and Curtis M. Grimm thony L. laquinto
1992 "Bridgingthe boundary:Exter- 1992 "Action characteristics as pre- 1989 "Inertia and creeping rational-
nal activityand performancein dictors of competitive re- ity in strategic decision pro-
organizationalteams." Admin- sponses." Management Sci- cesses." Academy of Man-
istrativeScience Quarterly, ence, 38: 439-455. agement Journal, 32: 516-
37: 634-665. Cyert, Richard M., and James G.
542.
Bantel, Karen,and Susan E. Jack- March Gersick, Connie J. G.
1989 "Marking time: Predictable
son 1963 A Behavioral Theory of the
1989 "Topmanagementand inno- Firm. Englewood Cliffs, NJ: transitions in task groups."
Academy of Management
vations in banking:Does the Prentice-Hall.
compositionof the top team Journal, 32: 274-309.
D'Aveni, Richard A.
make a difference?"Strategic 1994 Hypercompetition. New York: Ghemawat, Pankaj
ManagementJournal,10: Free Press.
1991 Commitment: The Dynamic of
107-124. Strategy. New York: Wiley.
Eisenhardt, Kathleen M.
Barbosa, Ricardo R. 1989 "Speed and strategic choice: Grimm, Curtis M., and Ken G.
1985 "Innovationin a matureindus- How managers accelerate de- Smith
try." Unpublisheddoctoraldis- cision making." California
1991 "Management and organiza-
sertation,ColumbiaUniversity. Management Review, 32:
tional change: A note on the
railroad industry." Strategic
Chaganti, Rajeswararao,and 533-576.
Management Journal, 12:
Rakesh Sambharya Eisenhardt, Kathleen M., and L. J. 557-562.
1987 "Strategicorientationand Bourgeois, III
characteristicsof upperman- 1988 "Politics of strategic decision Hambrick, Donald C.
agement." StrategicManage- making in high velocity envi-
1994 "Top management groups: A
ment Journal,8: 393-401. ronments: Toward a mid-
conceptual integration and
reconsideration of the 'team'
Chen, Ming-Jer, and Donald C. range theory." Academy of
label." In B. M. Staw and L. L.
Hambrick Management Journal, 31:
Cummings (eds.), Research in
1995 "Speed, stealth, and selective 737-770.
Organizational Behavior, 16:
attack: How small firms differ Eisenhardt, Kathleen M., and 171-214. Greenwich, CT: JAI
from largefirms in competi- Claudia B. Schoonhoven
tive behavior."Academy of Press.
1990 "Organizationalgrowth: Link-
ManagementJournal,38: ing founding team, strategy, Hambrick, Donald C., Ian C. Mac-
453-482. environment, and growth Millan, and Diana L. Day
Chen, Ming-Jer, and Ian C. Mac- 1982 "Strategic attributes and per-
among U.S. semi-conductor
formance in the BCG ma-
Millan ventures, 1978-1988." Admin-
trix-A PIMS-based analysis
1992 "Nonresponseand delayed istrative Science Quarterly,
of industrial product busi-
response to competitive 35: 504-529.
nesses." Academy of Man-
moves: The roles of competi- Finkelstein, Sydney, and Donald
tor dependence and action agement Journal, 25: 51 0-
C. Hambrick 531.
Academyof
irreversibility." 1990 "Top management team ten-
ManagementJournal,35: ure and organizational out- Hambrick, Donald C., and Phyllis
359-370. comes: The moderating role
A. Mason
of managerial discretion." Ad- 1984 "Upper echelon: The organiza-
tion as a reflection of its top
ministrative Science Quarterly,
35: 484-503. managers." Academy of Man-
agement Review, 9: 193-206.

Team Heterogeneity
Harrigan, Katherine R. MacMillan, Ian C., Mary L. McCaf- O'Reilly, Charles A., III,and Sylvia
1982 "Exit decisions in mature in- fery, and Gilles Van Wijk Flatt
dustries." Academy of Man- 1985 "Competitors' responses to 1989 "Executive team demography,
agement Journal, 25: 707- easily imitated new prod- organizational innovation, and
732. ucts-Exploring commercial firm performance." Working
banking product introduc- paper, University of California,
Hoffman, L. Richard, and Norman Berkeley.
tions." Strategic Management
R. F. Maier
Journal, 6: 75-86. O'Reilly, Charles A., Ill, Richard C.
1961 "Quality and acceptance of
problem solutions by mem- March, James G., and Herbert A. Snyder, and Joan N. Boothe
bers of homogeneous and Simon 1993 "Executive team demography
heterogeneous groups." Jour- 1958 Organizations. New York: and organizational change." In
nal of Abnormal and Social Wiley. G. P. Huber and W. H. Glick
Psychology, 62: 401-407. (eds.), Organizational Change
Michel, John G., and Donald C.
and Redesign: 147-175. New
Jacobson, Robert Hambrick
York: Oxford University Press.
1992 "The 'Austrian' school of 1992 "Diversification posture and
strategy." Academy of Man- top management characteris- Pindyck, Robert S., and Daniel L.
agement Review, 17: 782- tics." Academy of Manage- Rubinfeld
807. ment Journal, 35: 9-37. 1991 Econometric Models and Eco-
nomic Forecasts. New York:
Jackson, Susan E. Miller, Danny, and Ming-Jer Chen
McGraw-Hill.
1992 "Consequences of group 1994 "Sources and consequences
composition for the interper- of competitive inertia: A study Porter, Michael E.
sonal dynamics of strategic of the U.S. airline industry." 1980 Competitive Strategy: Tech-
issue processing. In P. Administrative Science Quar- niques for Analyzing Industries
Shrivastava, A. Huff, and J. terly, 39: 1-23. and Competitors. New York:
Dutton (eds.), Advances in Free Press.
1996a "The simplicity of competi-
Strategic Management, 8: tive repertoires: An empirical Prescott, John E., and Daniel C.
345-382. Greenwich, CT: JAI analysis." Strategic Manage- Smith
Press. ment Journal, 17: 419-439. 1987 "A project-based approach to
Jackson, Susan E., Joan F. Brett, competitive analysis." Strate-
1996b "Nonconformity in competi-
Valerie I. Sessa, Dawn M. Cooper, gic Management Journal, 8:
tive repertoires: A sociologi-
J. A. Julin, and K. Peyronnin 411-423.
cal view of markets." Social
1991 "Some differences make a Forces, 74: 1209-1234. Pugh, S. Douglas, and David J.
difference: Individualdissimi- Hickson
larity and group heterogeneity Miller, Danny, and Peter H. Frie-
1972 "Causal inference and the As-
as correlates of recruitment, sen
ton studies." Administrative
promotions, and turnover." 1977 "Strategy-making in context:
Science Quarterly, 17: 273-
Journal of Applied Psychology, Ten empirical archetypes."
276.
76: 675-689. Journal of Management Stud-
ies, 14: 253-280. Schumpeter, Joseph A.
Janis, Irving L. 1950 Capitalism, Socialism, and
1972 Victims of Groupthink. Bos- Mintzberg, Henry, Duru Raising-
Democracy, 3rd ed. New
ton: Houghton-Mifflin. hani, and Andre Thdoret
York: Harper.
1976 "The structure of 'unstruc-
Jauch, Lawrence R., Richard N. tured' decision processes." Smart, John C., and Ernest T. Pas-
Osborn, and Thomas N. Martin Administrative Science Quar- carella
1980 "Structured content analysis terly, 21: 246-275. 1986 "Self-concept development
of cases: A complementary and educational degree attain-
method for organizational re- Mischel, Walter
ment." Higher Education, 15:
search." Academy of Manage- 1968 Personality and Assessment.
3-15.
ment Review, 5: 517-526. New York: Wiley.
Smith, Ken G., and Curtis M.
Kmenta, Jan 1977 "The interaction of person and
Grimm
1986 Elements of Econometrics, situation." In D. Magnusson
1991 "A communication-information
2nd ed. New York: Macmillan. and N. S. Endler (eds.), Per-
model of competitive re-
sonality at the Crossroads:
Labich, Kenneth sponse timing." Journal of
Current Issues in Interactional
1990 "American takes on the Management, 17: 5-23.
Psychology. Hillsdale, NJ: Erl-
world." Fortune, 122: 40-48. baum. Smith, Ken G., Curtis M. Grimm,
Levine, Michael E. and Martin J. Gannon
Murray, Alan I.
1987 "Airline competition in deregu- 1992 Dynamics of Competitive
1989 "Top management group het-
lated markets: Theory, firm Strategy. Newbury Park, CA:
erogeneity and firm perfor-
strategy and public policy." Sage.
mance." Strategic Manage-
Yale Journal on Regulation, 4: ment Journal, 10: 125-141. Smith, Ken G., Curtis M. Grimm,
393-494. Martin J. Gannon, and Ming-Jer
O'Reilly, Charles A., III,David F.
MacMillan, Ian C. Chen
Caldwell, and William P. Barnett
1982 "Seizing competitive initia- 1991 "Organizational information
1989 "Work group demography,
tive." Journal of Business processing, competitive re-
social integration, and turn-
Strategy, 2: 43-57. sponses, and performance in
over." Administrative Science
the U.S. domestic airline in-
Quarterly, 34: 21-37.
dustry." Academy of Manage-
ment Journal, 34: 60-85.

Smith, Ken G., Ken A. Smith, J. D. Thomas, Anisya S., Robert J. Walsh, James P.
Olian, and H. P. Sims, Jr. Litschert, and Kannan Ra- 1988 "Topmanagementturnover
1994 "Top management team de- maswamy followingmergers and acquisi-
mography and process: The 1991 "The performance impact of tions." StrategicManagement
role of social integration and strategy-manager coalign- Journal,9: 173-183.
communication." Administra- ment: An empirical examina- Wiersema, Margarethe F., and
tive Science Quarterly, 39: tion." Strategic Management KarenA. Bantel
412-438. Journal, 12: 509-522. 1992 "Topmanagementteam de-
Spender, J. C. Virany, Beverly, Michael Tush- mographyand corporatestra-
1989 Industry Recipes: The Nature man, and Elaine Romanelli tegic change." Academy of
and Source of Managerial 1992 "Executive succession and ManagementJournal,35: 91-
Judgment. Cambridge: Basil organization outcomes in tur- 121.
Blackwell. bulent environments: An orga- Zenger, Todd R., and BarbaraS.
nization learning approach." Lawrence
Stalk, George, Jr.
Organization Science, 3: 72- 1989 "Organizational demography:
1988 "Time-The next source of
91. The differentialeffects of age
competitive advantage." Har-
vard Business Review, 88: Wagner, W. Gary, Jeffrey Pfeffer, and tenure distributionson
41-51. and Charles A. O'Reilly, Ill. technicalcommunication."
1984 "Organizationaldemography Academyof Management
Steiner, Ivan D.
and turnover in top-manage- Journal,32: 353-376.
1972 Group Process and Productiv-
ment groups." Administrative
ity. New York: Academic
Science Quarterly, 29: 74-92.
Press.
APPENDIX: Coding Categories for Executives' Backgrounds

Functional Background
1. CEO (Chief Executive Officer)
2. COO (Chief Operations Officer)
3. Finance/Treasurer
4. Planning
5. Personnel
6. Public Affairs
7. General Counsel/Secretary
8. Operations/Field Service
9. Marketing/Sales/Customer Service
10. Information Systems
11. International
12. Maintenance/Field Service
13. General Management
14. Other Corporate Staff
15. Accounting/Controller
16. Other
Educational Background
1. Engineering
2. Science
3. Business Administration
4. Economics
5. LiberalArts
6. Law (LL.B./J.D.)
7. Business (other than administration, e.g., accounting, finance)
8. Other


Hambrick 1996

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hambrick 1996

Uploaded by

Copyright:

Available Formats

The Influence of Top Management Team Heterogeneity on Firms' Competitive Moves

Author(s): Donald C. Hambrick, Theresa Seung Cho and Ming-Jer Chen

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

contended that competitionis fluidand that a firm advances

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

Other studies have found negative effects from top team

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

of initiative.In the context of competitive interplay,actions-

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

manifested in a propensityto launchcompetitive actions

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

homogeneity is better for routinetasks. Anotherdistinction,

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

the action. After identifyingthe airportsaffected by a firm's

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

where H is the heterogeneity measure and p the percentage

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

undergraduatedisciplines.We then calculatedthe Herfindal-

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

aB) a) I'* 00 CN CY)O i 00 CD 'It C-4 C-4 C-4 LO 'Iolt - O - M 01

m wm 00 M 00 CY) LO LO CY) 1 > d CN l CY) LO CD r- CB1N

CD-) ) C C 00C- COC 0 0 -

NC N C-4> CY) Q)C ~~~~~~~f 0 O

674/ASQ, December 1996

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

Functionalheterogeneity .122 .063--- .114-- .076" - .010---

N 156 153 642 642 603

Propositions2a-2c linkingtop management team heteroge-

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

Propositions5a and 5b, on response magnitude,were par-

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

port for these ideas. Allthree types of heterogeneity were

The results reveal a clear pattern:All three types of hetero-

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

Top Management Team Heterogeneity and

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

team homogeneity. It is also possible that after the period

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

682/ASQ, December 1996

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

683/ASQ, December 1996

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

APPENDIX: Coding Categories for Executives' Backgrounds

684/ASQ, December 1996

This content downloaded from 83.99.157.100 on Tue, 17 Jun 2014 08:50:32 AM

You might also like