Beruflich Dokumente
Kultur Dokumente
Quarterly http://asq.sagepub.com/
Published by:
http://www.sagepublications.com
On behalf of:
Additional services and information for Administrative Science Quarterly can be found at:
Subscriptions: http://asq.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations: http://asq.sagepub.com/content/47/1/70.refs.html
What is This?
METHOD
Data
I investigatedthe effects of corporateculturestrength on the
reliabilityof firm performanceusing Kotterand Heskett's
(1992) data on the strength of corporatecultureamong a
sample of large, publiclytradedfirms in 18 markets. Kotter
and Heskett (1992) began with 21 markets defined according
to their own criteria.In general, these markets are analogous
to the marketcategories used in Fortunemagazine. Missing
data problems led Burtet al. (1994) to reduce the number of
industriesto 19 by eliminatingthe life insuranceindustryand
combiningsavings and loans and commercialbanking.Inthe
currentanalyses, missing data on the bankingindustryresult-
ed in the furtherexclusion of that industryfrom the analyses.
Kotterand Heskett asked the top six officers in the firms
selected for the study to complete a short, mailed question-
naire.The respondents were asked to assess the strength of
the corporateculture in each of the other sampled firms that
were in the same industryas their own firm. Respondents
were asked to assess the strength of corporateculture in the
late 1970s and early 1980s by judgingthe degree to which
managers at a firmwere influenced in their decision making
by a strong corporateculture (Kotterand Heskett, 1992: 161).
The survey providedrespondents with three indicatorsof a
strong corporateculture:(1) managers in the firmcommonly
speak of their company's style or way of doing things; (2) the
firm has made its values known througha creed or credo and
has made a serious attempt to get managers to follow them;
and (3) the firm has been managed accordingto long-stand-
ing policies and practices other than those just of the current
chief executive officer. Respondents were asked to rate each
77/ASQ, March 2002
Downloaded from asq.sagepub.com at St Petersburg State University on December 19, 2013
firmon a scale of one to five, and an average score across
respondents was computed for each firm.While Kotterand
Heskett (1992) used an inverse coding of culturestrength,
with 1 indicatinga strong culture, I have reversed their cod-
ing to ease interpretation.
A strong advantageof this measurement strategy is that it
measures culturestrength across a very broadrange of firms
and competitive contexts without requiringextensive culture
surveys withinfirms. Data on a varietyof industriesis critical
to testing the claim that the relationshipbetween culture
strength and reliabilityis contingent on industryvolatility.Kot-
ter and Heskett's measurement strategy leads a firmto be
characterizedas havinga strong cultureif other actors in its
industryassociate the firmwith a uniqueand common way
of doing things, relativeto other firms in the industry.More-
over, this distinctivebehaviormust be codifiedand have per-
sisted over time. This culturestrength variabledoes not
directlymeasure the extent to which there is consensus
within the firm, however, and an externalassessment of cul-
ture strength is suboptimalin certainrespects. It is possible,
for example, that some strong-culturefirms identifiedin the
sample are highlyfragmented but manage to projectan aura
of cohesion and consensus. It is importantto keep in mind,
however, that Kotterand Heskett's questionnaireasked
respondents to characterizefirms accordingto differentspe-
cific behavioralcharacteristics,not simplyto generate a glob-
al culturerankingthat might be more subject to image
manipulation.It seems less likelythat weak-culturefirms will
be able to exhibitbehavioralconsistency over time.
Furthermore,surveyingorganizationalmembers about their
firm'sculturemay be problematicwhen studyingvariabilityin
performance.As Weick (1985: 386) suggested, people may
be more likelyto attend to culturewhen their dailyroutines
breakdown and they are presented with unfamiliarsitua-
tions. The reliabilityof firm performancemay thereby affect
the measurement of culturestrength. Whilethis type of
response effect may affect externalevaluatorsas well, it
arguablyhas the greatest impacton internalinformants.
Kotterand Heskett (1992) checked the validityof their mea-
sure in several ways. Most importantly,they conducted inter-
views with managers of a selected subsample of firms in
which they asked respondents a series of questions about
the strength of corporateculture.The resultingscores corre-
lated well with the externalmeasures. Ultimately,however,
their measure assumes that firms with widely shared and
deeply held norms and values will exhibitthese externally
observable characteristicsand that firms characterizedby dis-
sension will be unlikelyto exhibitthe same characteristics.
An additionalconcern, as Burtet al. (1994) noted, is that the
culturescores generated by Kotterand Heskett's design
might reflect a response effect. Respondents may make
inferences about a firm'sstrength of corporateculturebased
on its performanceor size. Firmsthat are more visible or
salient to a respondent may receive higherculturescores. In
fact, a regression of culturestrength on a firm'saverage mar-
ket capitalizationbetween 1979 and 1984 (controllingfor
78/ASQ, March2002
Downloaded from asq.sagepub.com at St Petersburg State University on December 19, 2013
Corporate Culture
Table1
Descriptive Statistics*
Variable Mean a 1 2 3 4 5 6
1. ROI .09 .08
2. Operatingcash flow 1342.27 2895.10 .02
3. Debt-to-assetratio .41 .18 -.59' .03
4. Operatingleverage .42 .17 -.26* .34* .19'
5. Culturestrength 3.29 .79 .30' .29' -.40' .06
6. Log marketcapitalization 7.22 1.42 .25' .58' -.42' .21 .38'
7. Log marketconstraint -1.49 .72 -.26' -.21 .33' -.11' -.14' -.43'
p< .05.
* Culturestrength ranges from 1 to 5, with higherscores indicatinga strongercorporateculture.
=
ri Oi + 3rm + ?
Table 2
=
Yj JLi + OiEi
ni = E(y,)= P'X
ci = Var(y,)= exp(y'Zi)
Table3
Multiplicative Heteroscedasticity Models of FirmPerformance Measures, 1979-1984*
ROI Operatingcash flow
Variable (1) (2) (3) (4) (5) (6)
Mean:
Debt-to-assetratio -.170- -.170" -.164- -3.301 2.200 45.732
(.012) (.012) (.012) (31.645) (31.496) (28.424)
Operatingleverage -.082# -.080" -.072" 301.249" -42.230 -59.631
(.014) (.014) (.014) (7.056) (46.591) (37.732)
Corporateculturestrength .006" .006" .017- 14.575 -.563 19.495
(.002) (.002) (.010) (1.559) (5.887) (16.057)
Log marketcapitalization -1.E-04 -2.E-04 14.009 15.188t
(.001) (.001) (6.902) (5.883)
Culturestrength x Log market .006- 5.014
constraint (.002) (11.072)
Log variance:
Corporateculturestrength -.41 8 -.437- -.683- .940- -.277- -1.790"
(.079) (.086) (.167) (.096) (.116) (.198)
Log marketcapitalization .073 .086 1.454- 1.485-
(.062) (.062) (.085) (.086)
Culturestrengthx Log market -.182 -.927-
constraint (.111) (.110)
X2 1038 1019 1031 3339 3589 3646
D.f. 38 40 42 39 41 43
Firm-yearspells 806 800 800 703 703 703
-
*p < .05; p < .01; two-sided tests.
* Models includeindustrydummies in the predictionequationsfor both the mean and
log variance;thus, all covariates
are relativeto the marketaverages. The model for operatingcash flow includesa controlfor the operatingcash flow
level in 1978. Standarderrorsare in parentheses.
82/ASQ, March2002
Downloaded from asq.sagepub.com at St Petersburg State University on December 19, 2013
Corporate Culture
Table5
Contingent Effects of Culture Strength on FirmPerformance, 1979-1984*
ROI Operatingcash flow
Variable (1) (2) (3) (4)
Mean:
Debt-to-asset ratio -.1 69" -.1 66" 71.801 83.023"
(.012) (.012) (32.515) (28.749)
Operatingleverage -.080" -.073" -81.142 -73.242*
(.014) (.014) (41.260) (34.926)
Corporateculturestrength .003 .021 51.027 66.070"
(.006) (.009) (29.767) (29.162)
Log marketcapitalization -.001 -4.E-04 2.539" 18.458-
(.001) (.001) (6.495) (5.509)
Industryvolatility -.008 -.009 36.939 31.107
(.010) (.010) (28.124) (25.016)
Culturestrengthx Industryvolatility -.003 -.004 -54.975 -49.109
(.006) (.007) (35.478) (33.846)
Culturestrengthx Log marketconstraint .006" 8.530
(.002) (9.606)
Log variance:
Corporateculturestrength -1.816" -1.994" -2.382" -3.225W
(.310) (.344) (.364) (.316)
Log marketcapitalization -.005 .021 1.538" 1.532"
(.066) (.066) (.090) (.090)
Industryvolatility .419 .414 -.841 -1.134-
(.411) (.409) (.464) (.465)
Culturestrengthx Industryvolatility 1.428" 1.361 - 1.951 - 1.572"
(.307) (.307) (.342) (.308)
Culturestrengthx Log marketconstraint -.180 -.846"
(.112) (.105)
84/ASQ, March2002
Downloaded from asq.sagepub.com at St Petersburg State University on December 19, 2013
Corporate Culture
1__
0- /
0cJ /J
-1- s ,'
-J
o r- /
r 2- - /.
0 .5 1 1.5 2
Industry Volatility
90/ASQ, March2002
Downloaded from asq.sagepub.com at St Petersburg State University on December 19, 2013
Corporate Culture