Drawing on survey and performance data for 34 corporations, this study shows thatorganizations that have participative corporate cultures and well-organizedworkplaces have better performance records than those that do not.
BringingCorporate Culture To the Bottom Line
Daniel R. Denison
ST he impact of corporate culture on the design core identity of an organization. A “strong”and management of organizations is a con- culture that encourages the participation andstant theme in contemporary writing about involvement of an organization’s membersAmerican business. Most authors agree that appears to be one of its most important“corporate culture” refers to the set of values, 英國dissertation網assets.beliefs, and behavior patterns that form the The impact of corporate culture has 5been cited as an explanation for the differencesin productivity among Americanfirms, and the differences in productivity betweenAmerican and Japanese companies.Superior Japanese productivity consistentlyhas been attributed, in part, to better organizationof work, consensus decision making,and an elusive quality called the effectivemanagement of human resources. The “soft”side of management has seldom received somuch attention.
Despite this attention, there is littlesolid evidence about the impact of an organization’sculture on performance. The evidencethat does exist is seldom presented ina form that is convincing to managers andexecutives and, therefore, these humanresources issues often remain on the backburner.
BACKGROUND AND RESEARCH METHODOLOGY
The study described in this paper directly addressesthe problem of “no evidence” andasks these questions: Do firms that organizetheir work well and involve employees in decisionmaking really perform any better thanfirms that do not? If so, how much better,and under what conditions?
Using survey data as an indicationof cultural managerial style and Standardand Poor’s financial ratios as indicators ofperformance, this research compares a set of34 large American firms in order to begin totest the relationship between corporate cultureand performance. The results, presentedin terms of return on investment and otherfinancial indicators, indicate that companieswith a participative culture reap a return oninvestment (ROI) that averages nearly twiceas high as those in firms with less efficientcultures. The data presented here providehard evidence that the cultural and behavioralaspects of organizations are intimatelylinked to both short-term performanceand long-term survival.Measuring Corporate CultureThe measurement of something as complexand amorphous as an organization’s culturehas been the subject of much debate. Someargue that each culture is unique and must beintuitively “sensed” rather than measured;
others argue that the best way to uncoverculture is through ethnographic studies thatanalyze the stories and accounts of eventsmaking up the folklore of every organization.Recent special issues of Organizational#p#分頁標題#e#
Dynamics and the Administrative Science
Quarterly have focused exclusively on organizational
culture and have elaborated onthe new approaches that have been taken tostudy this phenomenon. The study describedhere takes still a different approach in ineasuring
culture, but relies on an establishedresearch technique - the survey index.Daniel R. Denison, project director and assistantresearch scientist at the University ofMichigan’s institute for Social Research, teachesin Michigan’s Organizational Psychology program.He received his Ph.D. from Michigan in1982. His research and consulting have centeredon a number of topics, including organizationalculture and effectiveness, the growth and developmentof new organizations, applying thework-team concept in organizational design,the impact of technological innovation, and theuse of multivariate methods in theory testing.Dr. Denison is currently studying the relutionshipbetween the quality of work life andthe differences in effectiveness between unitswithin a large manufacturing corporation. This
research examines intraorganizationul differencesin much the same way that interorgunizutionuldifferences are studied in this article.The complete project upon which this article
is bused is described in the author’s forthcomingbook, Corporate Culture and Organizational
Effectiveness: A Behavioral Approach toFinancial Performance (Wiley-Interscience,1985).
ganizations in the same way. The results thenprovide a basis for comparison and generalization.
For example, a set of items combinedinto an index of communication might indicatethat one organization had better communicationthan a second organization thatscored lower on the index. A finding that
firms with better communication also werebetter performers then could be taken as evidenceof a close relationship between communicationand performance.
One disadvantage of the surveymethod is that there is no safeguard againstovergeneralization. Comparing decisionmakingpractices in organizations, for example,can sometimes be like comparing apples
and oranges. Assigning an organization (or adivision of an organization) a single score on*decision-making practices implies that all decisionswithin that unit are made in the same
way. It also implies that participative decision-making practices are similar and willhave the same effects in a sales or bankingorganization as in a manufacturing company.
If this basic comparison is not justified,Using a survey technique to studyorganizational culture has both advantagesand disadvantages. The key strength is that
the same method can be applied to many or- 7it can result in other unwarranted comparisonsand generalizations.
Both the comparative surveymethods used in this study and the ethnographicmethods emphasizing the unique
character of each organizational setting sharea common focus: the set of values and beliefsthat lie at the core of an organization’s cultureand the practices and patterns that stem#p#分頁標題#e#
from, and reinforce, those basic values.These issues reflect a recurring theme in organizational
studies over the past 30 years:the central importance of values and beliefsto organizational life.
The data used in this study werebased on individual perceptions of organizationalpractices and conditions; these perceptionswere used to characterize the culture ofeach organization. The data on 34 companieswere drawn from the Survey of Organizationsarchive at the University of Michigan’sInstitute for Social Research (ISR).Measuring Organizational Performance
The method of measuring an organization’sperformance also is a controversial issue.The approach taken by this study is a simpleone, selected for its relevance to managersand executives as well as to researchers andacademics. A number of financial ratios, includingreturn on investment, equity, andsales, were computed for all 34 firms thatwere in the ISR archive and also were listedon either the New York or the AmericanStock Exchange. The financial data weretaken from Standard and Poor’s statisticalservice, COMPUSTAT. The ratios for eachcompany were compared with those of itscompetitors 英國留學生dissertationwithin each industry to producea standardized score representing the firm’scompetitive standing.Although financial ratios are not
the only, or even the best, indicators of organizational
performance, effectiveness does
imply that an organization can successfully
meet the demands of a broad set of stakeholders:
investors, shareholders, employees,
customers, suppliers, and so forth. Of course,
complete reliance on financial indicators of
business performance often can bias a measure
of effectiveness toward particular stakeholders.
For example, risk-adjusted earnings
per share is a favorite measure of business
performance for market analysts, but shows
a clear bias toward shareholders and investors.
This study’s business-performance
measures, emphasizing the organization’s
ability to generate income, are in keeping
with a definition of effectiveness that focuses
on an organizations capacity to acquire resources
from its environment. This measure
of success generally reflects the interests of all
stakeholders, even though the strategies for
acquiring resources often involve clear tradeoffs
among stakeholders. Future research
may study corporate culture’s impact on effectiveness
with a more comprehensive set of
measures.
The Survey Data
The Survey of Organizations, a 125-item
standardized questionnaire developed over
the past 15 years at ISR, operationalized the
theories of Rensis Likert, one of the original
proponents of participative management.
Since its genesis, the instrument has been
periodically updated to incorporate new developments#p#分頁標題#e#
within the field of organizational
behavior. The archive at the University of
Michigan now incorporates survey data from
over 300 organizations. The instrument’s
validated scales on organizational climate,
work design, leadership, group functioning,
and satisfaction focus on respondents’ perceptions
about the way their organization is
managed, rather than on their own attitudes,
so that a clear, diagnostic picture of an organization
may emerge from the survey data.
The survey design presumes that
certain social processes and relationships are
common to all organizations and have a consistent
correlation with performance and
effectiveness. This controversial set of assumptions
allows organizations to be compared
on social, behavioral, and cultural
dimensions in a way that case studies and
anecdotes cannot be compared. In principle,
other dimensions of culture also can be
measured and compared in this same manner.
Their relationship to performance at this
point remains an empirical question.
This study was based on the perceptions
of 43,747 respondents in 6,671 work
groups in 34 companies that were included
both in the Survey of Organizations archive
and Standard and Poor’s COMPUSTAT listing.
These 34 companies represent 25 different
industries as defined by Standard and
Poor’s four-digit industry code. A listing of
the industries from which this sample was
drawn is presented in Exhibit 5.
Each individual’s responses to the
survey items were averaged, by topic, into 22
indexes in the areas of organizational climate,
leadership, peer relations, group
process, work design, and satisfaction. Each
of these individuals’ index scores were then
averaged with the other members of the same
work group to get a group score. Finally, all
work group scores were averaged to get a
score for the entire organization on each
index.
The companies in this sample were
“self-selected”; they are client organizations
that voluntarily chose to use the survey some
time during 1966-1981. Since the companies
also selected the divisions that would be included
in the survey sample, not all members
of all organizations are represented in the
sample.
Performance data in all cases referred
to the entire corporation as listed on
the New York or American Stock Exchange.
Several tests were done to make certain that
a faulty match did not artificially account for
the results of the study. When the poorest
matches (e.g., survey data from one small
division of a large corporation matched with
9
10
its entire performance data) were excluded,
the observed relationship between the survey#p#分頁標題#e#
data and the performance data not only remained,
but increased.
The Financial Performance Data
A number of COMPUSTAT’s financial indicators
initially were examined before settling
on those reported here: income/investment
and income/sales ratios. Income/equity ratios
yield similar results but in several instances
are more difficult to predict, given
their sensitivity to debt-financing versus
equity-financing strategies.
Since many theories predict that
the relationships examined in this study
would vary greatly in accordance with the
business environment and industry, some
means of comparing performance within the
same industry category also was necessary.
Still it was impossible to do any withinindustry
analysis, since no more than two
firms within any particular industry were
examined.
The COMPUSTAT data, however,
are available for many firms within each industry
category and thus provide a basis for
standardization of the financial data for
those firms that were in the study sample.
The financial data of each firm for each year
were compared with all firms listed by COMPUSTAT
within the same industry in order
to compute a standardized score. These standardized
scores were then converted to a percentile
score (scaled from 1 to 100) for
the analyses presented here. The industries
represented in the study, the number of firms
in each industry, and the number of firms
used for comparison are presented in Exhibit
5.
This standardization procedure accomplishes
two important objectives: It allows
for a test of the impact of the behavioral
and cultural variables on performance against
competitors rather than according to an absolute
measure of performance, and it eliminates
the effects of the overall economic
climate.
Return on investment and return on
sales, and their standardized equivalents,
were computed for each firm for the five
years following the year in which the survey
data were gathered. Not all survey data were
gathered in the same year, but subsequent
performance data are included as year 0
through year +5 for all companies. In this
way, characteristics of the companies’
management systems and practices, measured
at one point in time, were used to
predict performance for the five years that
followed.
THE RESULTS: A FIVE-YEAR COMPARISON
The original results of this study were compiled
as correlations between the survey
measures and performance over time. A
more graphic way to look at the results of
this research, however, is to divide the sample
of organizations in half and compare the
organizations that are above average on the
survey indexes with those that are below#p#分頁標題#e#
average. The results are presented here for
two of the indexes that had the largest impact
on performance: the organization-of-work
index and the decision-making practices
index.
Each of these indexes makes reference
to a behavioral or cultural feature of an
entire organization. For example, one of the
items in the decision-making practices index
asks: “People at all levels of an organization
usually have know-how that could be of use
to decision makers. To what extent is information
widely shared in this organization so
that those who make decisions have access to
such knowledge?” In general, measures like
these that make reference to the management
system of an entire organization are the best
predictors of long-term performance. Indexes
that refer to leadership styles or group functioning,
rather than to the system as a whole,
are generally better as predictors of short- to
medium-term performance. Systemwide cultural
characteristics seem to have the most
enduring impact on performance.
The organization-of-work index is
a composite of four survey items that reflect
the degree to which work is sensibly organized,
work methods are adapted to
changing conditions, decisions are made at
appropriate levels, and the goals of the organization
are perceived by the individual as
clear and reasonable. Each respondent’s answers
to the four items are compiled as an index
that is then averaged for all members of
his or her work group. The average of all
work groups in an organization determines
the index score for that organization as a
whole.
The decision-making practices index
is a two-item measure indicating the
degree of involvement that individuals have
in the decisions that affect them, and the extent
to which information is shared across
levels of an organization in a way that brings
the best information possible to decision
makers.
The impact of these two indexes is
presented in terms of two indicators of performance
and their standardized equivalents:
(1) The income/investment ratio compares
income (after all expenses, income taxes, and
minority interests, but before provisions for
common and/or preferred dividends) with
total investment (long-term debt, preferred
stock, minority interest, and common equity).
(2) The income/sales ratio compares income
with net sales (gross sales reduced by
cash discounts, trade discounts, returned
sales, and other allowances).
These two financial ratios reflect
quite different aspects of performance. Income/
investment is a measure of the effective
utilization of resources over time, while
the income/sales ratio is more of an indicator#p#分頁標題#e#
of operating efficiency. The standardized
ratios compare each company to competitors
in the same general industry in the same year
and then express performance compared
with that of competitors as a percentile score.
Exhibit 1 compares return-on-in-
“The survey design presumes fkaf certain
social processesa nd relationshipsa re
cornrnon to all organizations and have
a consisfenf correlation wifk
performancea nd effectivenes‘sI . 11
COMPARISON OF RETURN-ON-INVESTMENT RATES FOR COMPANIES HIGH AND Low ON THE
ORGANIZATION-OF-WORK INDEX (BY PERCENTAGE AND PERCENTILE FOR A FIVE-YEAR PERIOD)
Organization of Work and Return on Investment
0 +1
Years
+2 f3 f4
20%.
1.5%.
Return on Investment
Percentage 10 % .
O%-
highs
lows --w-w
Percentile Score by Industry
Years
f5
0 +1 f2 +3 +4 +5
,--
90 I
80
70
60
Return on Investment ,---- /’
Percentile
5. c@ --r, /
/
40 --. /
30 -\ -4’
20 highs
10 lows -w--w
0.
vestment rates for the companies that are ously, this study divided the sample of orhigh
and low on the organization-of-work ganizations in half. The “highs” in Exhibits 1
index and the corresponding percentile and 2 simply represent those companies
12 scores for these ratios. As mentioned previ- whose scores are above the J&firm average
on the organization-of-work index, while the
“lows” represent those that are below the 34-
firm average. These data clearly show that
those companies perceived as having a wellorganized
work environment as have a significantly
higher return on investment. The
highs, in fact, often have an ROI that is twice
as high as the lows. Most importantly, these
differences are not temporary-good work
organization in year 0 is not only a good indicator
of high performance today, but also
seems to predict a high return on investment
as far as three to five years into the future.
The same pattern also appears, but
with more stability, in the percentile scores:
the highs have a ranking within their industry
that is from 15 to 50 percentage points
higher than the lows’ rank within their industry.
This difference appears in all years, and
with the exception of the last year (+5) the
gap between the highs and lows appears to
widen over the years. (The percentile comparisons
also show that the sample of firms
for which Survey of Organizations data were
available typically performed slightly better
than their competitors. This raises a few
questions of interpretation, but doesn’t call
into question the basic underlying pattern.
The relationship between culture and performance#p#分頁標題#e#
may, in fact, have appeared stronger
if the study sample had included more firms
that performed poorly.)
Exhibit 2 contrasts the second performance
measure, return on sales, with the
organization-of-work index. The high versus
low contrast for this financial indicator of organizational
efficiency is even more striking
than the return on investment analysis. The
differences between the highs and the lows
are substantial in year 0 and grow consistently
wider during years +l through +5.
Both of these analyses show that
companies with a culture that encourages the
development of adaptable work methods
linking individuals to the goals of an organization
have a clear competitive advantage.
This advantage appears to be substantial
when expressed in terms of return on investment,
and seems to have an even stronger
impact when presented in terms of the efficiency
measure, return on sales. The organization-
of-work index captures the potential
of a company to efficiently reorganize and
adapt both in the present and in the future.
Exhibits 3 and 4 present the results
in a similar way for the second survey index,
decision-making practices. Exhibit 4 shows
the differences in return on investment between
those companies that rely on participative
decision-making practices and those
that do not.
These results show a very different
pattern from the findings regarding organization
of work. Large differences in return on
investment do not appear at years 0, +l, or
+2. The data for year +2, in fact, even seem
to indicate that firms that have more participative
decision-making practices have a
slightly lower return on investment. This absence
of a performance difference for years 0
through +2 dramatically reverses in years
+3 through +5; however, the highs outperform
the lows by a factor of two or three to
one.
The percentile comparisons show a
slightly different picture. When performance
relative to competitors is examined, there appears
to be a small initial advantage associated
with a participative culture that steadily
widens over the five-year period. Performance
relative to competitors steadily increases
for the highs in this sample, moving
roughly from the 60th percentile to the 80th
percentile over the five years that these firms
were studied.
The relationships between involve- 13
Exhibit 2
COMPARISON OF RETURN-ON-SALES RATES FOR COMPANIES HIGH AND Low ON THE
ORGANIZATION-OF-WORK INDEX (BY PERCENTAGE AND PERCENTILE FOR A FIVE-YEAR PERIOD)
20%
15%
Return on Sales
Percentage 10 %
0%
100
90
80
70
60
Return on Sales 5.
Percentile
40#p#分頁標題#e#
30
20
10
0
Organization of Work and Return on Sales
Years
0 +1 f2 +3 t4
highs -
lows -----a
Percentile Score by Industry
Years
highs -
lows -w--m
5
0 $1 +2 +3 +4 f5
ment in decision making and performance and one that takes some time to pay off. Takand
between organization of work and per- ing more time to involve managers, execuformance
are different. Participation in deci- tives, or employees in a decision may not
14 sion making appears to be an investment, always be the quickest way to make a deciExhibit
3
COMPARISON OF RETURN-ON-INVESTMENT RATES FOR COMPANIES HIGH AND Low ON THE
DECISION-MAKING PRACTICES INDEX (BY PERCENTAGE AND PERCENTILE FOR A FIVE-YEAR PERIOD)
Decision-Making Practices and Return on Investment
Years
0 +1 f2 +3 f4 +5
20%
15%
Return on Investment
Percentage 10 %
5%
0%
Return on investment
Percentile
100
90
80
70
60
50
40
30
20
10
0
highs -
lows -m-m
Percentile Score by Industry
Years
0 +1 +2 +3 i-4 +5
: --
--- -t-- - ---_ -x -. -\
-I
zccHhighs
-
lows ----
sion, but the evidence here and in other affected are involved in a decision before it is
studies suggests that it may lead to a better made.
decision. More importantly, implementation Exhibit 4, the final set of results,
is usually improved when those who will be compares return on sales for those firms that 15
Exhibit 4
COMPARISON OF RETURN-ON-SALES RATES FOR COMPANIES HIGH AND Low ON THE DECISION-MAKING
PRACTICES INDEX (BY PERCENTAGE AND PERCENTILE FOR A FIVE-YEAR PERIOD)
Decision-Making Practices and Return on Sales
0
20% -
fl
Years
+2 +3 +4 +5
t
15% -
Return on Sales
Percentage 10% -
highs -
lows ---w
Return on Sales
Percentile
Percentile Score by Industry
Years
,;;; ‘,l +,2 +,3 +p +5
50
40
30
20
10
-. -Y- ---
highs -
lows ----
are high and low on participation in decision the high- and low-participation companies
making. These results look very similar to are small, but they grow consistently wider
the return on investment analysis presented over the five-year period that was studied.
16 in Exhibit 3; the initial differences between The results presented in Exhibits 3
and 4 illustrate the impact that a participative
culture can have on performance against
competitors. Standardizing the performance
measures by industry makes the relationship
more, not less, apparent. Standardization
also resolves some of the confusion caused
by the absence of a positive relationship between#p#分頁標題#e#
participation in decision making and
performance in years 0 through +2 that appear
in the analysis of the unstandardized
performance measures in Exhibits 3 and 4.
The most interesting feature of this
study is that the same general pattern appears
when performance against competitors
is used in place of the simple performance
measures. Many interpretations of contingency
theory would predict that the relationship
should be diminished when performance
is standardized by industry. Contingency
theory predicts’that in stable environments,
participation should have an adverse effect
on performance, while in changing environments,
participation should have a positive
effect. Since characteristics of the business
environment vary by industry, combining
companies from a diverse set of industries
should diminish the observed performance
impact of participation. This effect does not
show up in the results. The implications of
this for the “contingency” versus “one best
way” debate are further addressed later on in
this article.
These results may come as a surprise
to those who think of corporate culture
or participatory decision making as being
too soft or too amorphous to have practical
implications. Nevertheless, this research
shows that performance differences do exist.
The impact on companies is substantial, and
this study suggests, as have many theorists,
that the management of an organization’s
culture (whether implicit or explicit) should
be one of the fundamental elements of a corporation’s
strategy for staying in business.
These findings also give an indication
of the potential that exists for monitoring
an organization’s management system
and culture and for predicting their impact
on performance in the future. As this capacity
for prediction develops, the potential for
diagnosing, directing, and managing an organization’s
culture will also become much
clearer. These initial results have focused on
two basic aspects of an organization’s culture
and have shown that each has a direct and
long-lasting effect on the financial performance
of a set of organizations. Soft measures
do, in fact, predict hard outcomes.
ADDITIONAL FINDINGS
This article has described a small, but highly
relevant part of a much larger study. The
results from only 2 of 22 survey indexes have
been presented in a simplified manner to
make the point as directly as possible. There
are many other interesting findings from the
study, some equally strong and some more
tentative. Some of these findings follow.
The unit of analysis. The survey
data included individuals’ attitudes toward#p#分頁標題#e#
their immediate work groups, their supervisors,
other groups with whom they interact,
and the organization as a whole. The
system-level attitudes were the best predictors
of performance, particularly in the
long run. Attitudes regarding respondents’
work groups or supervisors worked
moderately well as predictors of short- to
medium-term performance, but did not predict
longer-term performance.
These results seem to indicate that
(1) studies that attempt to predict performance
often “mismatch” units of analysis,
and often cannot pinpoint an outcome variable
that reflects true performance while still
allowing for comparison between organiza- 17
18
tions; and (2) a close match between the
units of analysis for independent and dependent
variables, along with an accurate and
comparable measure of performance seems
to be necessary to uncover the true relationship.
Furthermore, these results seem to
highlight the tremendous inertia in an organization’s
culture and management system;
it is the system-level characteristics that influence
performance far into the future.
The importance of high ideals. The
study included a limited amount of information
about preferred patterns of leadership
gleaned from questions that asked respondents
to rate their ideal supervisor on a number
of dimensions. In several cases, these
ratings of the ideal leader were better predictors
of performance than were the ratings of
actual supervisory behavior. The lowest performers
in this study were those organizations
in which ideals seemed unimportant or
unclear. These data seem to provide some
support for one of the most significant contributions
of the emerging cultural theory of
organizations: the importance of a symbolic
vision as a means of providing direction and
integration. The ideal or vision that an organization’s
members hold appears to be as
important to performance as the actual behavior
of the company’s employees.
Consistency, agreement, and conformity.
The data presented from the organization-
of-work and decision-making practices
indexes compared the level of responses
on those indexes with organizational performance.
A different way to conceptualize the
relationship between management practices
and performance is to look at the consistency
of responses across groups within each organization
and the level of agreement about
management practices throughout the organization.
Such a consistency measure is one
way of looking at the effects of conformity
and is a good way to operationalize the degree
to which an organization has a strong
culture.
The results of this part of the study#p#分頁標題#e#
are intriguing, even though they probably
raise more questions than they answer: High
consistency is associated with high current
performance and short-term performance,
but is associated with low long-term performance.
One interesting interpretation of this
finding is that consistency is an indication of
a system that is currently well coordinated
and integrated, and that currently performs
well. In the longer term, however, the lack of
variety connected with such a system limits
the organization’s ability to adapt to changes
in the environment. Such speculation is difficult
to prove or disprove with these data, but
clearly suggests areas for future study. The
interaction of a strong and participative culture
may well lead to the most favorable performance
conditions, but these ideas have
not yet been fully examined.
KEY QUESTIONS RAISED BY THIS STUDY
The findings of this study raise some key
questions about organizational theory and
how organizations are studied. Several of
these questions are addressed below.
Why Participation Works
The data presented here provide evidence
that participation works but little information
on why it works. The question is a complicated
one, but at least four processes seem
to be associated with high participation and
involvement.
1. A participative culture encourages a
higher degree of inclusion of the individual
in the work environment. Workers become
more than “hired hands” and develop a sense
of ownership over their efforts within the organization
and a pride in their contributions
to the workplace. An environment is created
that requires an organization’s members to
become psychologically involved and to become
aware of and concerned with the consequences
of their actions.
2. Coordination within a participative
culture becomes an a priori condition of
planning, problem solving, and decision
making, rather than an afterthought or an
element in some post hoc implementation
plan. The actors are informed and seek to
minimize the transaction costs associated
with conflicts while defining a course of action
that furthers their collective interests.
3. Participation fosters the long-term development
of responsible work habits on the
part of individual members. When participation
is a part of an organization’s culture, individuals
tend to develop a view of themselves
and the organization that extends
beyond their immediate job or working situation.
Over time, identification with organizational
goals develops.
4. Groups of people do a better job of
solving complex, multifaceted problems than
do individuals. This is particularly true if the#p#分頁標題#e#
interactions of groups members are structured
in an effective manner. A set of norms
that capitalize on this fact probably helps
some organizations to consistently make
better decisions.
Contingency Theory versus One Best Way
The results presented here could be taken as
a triumph of “one best way” theories over
theories that argue that the best form of
organization is dependent on the situation
and the business environment. In reality,
both points of view are probably correct,’
and the two ideas are not nearly so inconsistent
as the years of debate have made them
seem. It is more likely that there are some
universals and, at the same time, many factors
that are situationally dependent. Contingency
theory might even be viewed as a
two-factor, one best way theory that pairs
autocracy with stability and participation
with turbulence.
The more relevant questions to ask
would seem to be, What is the relative impact
of the “main effect” (participation) and
the “contingent effect” (the business environment)?
and Under what conditions does the
relative impact change? The data presented
here appear to support the argument that
participation seems to have a positive effect
in a broad range of situations, but do not
“Contingency theory nz&ht even be viewed
as a two-factor, one bed way theory fhaf
pairs autocmcy wifh stability and
pmticipdon with turbulence‘.I 19
20
rule out the possibility that the positive effect
may be much greater in those situations in
which the business environment changes
rapidly. Overall economic stability and more
global rates of change also may alter the relative
impact of contingent effects.
Corporate Culture and Participation
The results and discussion presented here are
not meant to equate organizational culture
and participation. Participation is only one
element of corporate culture that may have a
clear impact on organizational performance.
Future research will soon expand the list of
universal or situationally-specific cultural
aspects of organizations that influence performance.
One consistent theme in the culture
literature concerns the impact of a strong culture
on organizational performance. The
hypothesis is that an organization with a
high level of shared meaning, a common vision,
a “clanlike” attitude toward members,
and a high level of normative integration will
perform well. Some authors have argued that
this theory has universal application, while
others have argued that the culture of an organization,
in addition to having these
characteristics, must also fit the business environment
.#p#分頁標題#e#
The strong-culture hypothesis bears
an interesting relationship to the participation
hypothesis; both emphasize inclusion.
The culture literature has tended to emphasize
shared meaning, while the participation
literature has focused more on managerial
practices and style; however, both have
stressed the central importance of the psychology
of inclusion.
The content of an organization’s
culture, however, need not include participation.
At this point the strong-culture hypothesis
and the participation hypothesis
diverge and make quite different predictions.
An organization that is rigid, autocratic,
conflict-ridden, and individualistic, could,
under the strong-culture hypothesis, still perform
very well as long as it fit a stable business
environment and met the criteria for
inclusion, meaning, vision, and integration.
In that case the participation hypothesis
would predict poor performance, while the
strong-culture hypothesis would predict the
opposite. It is hoped that future comparative
research will resolve this paradox and other
key issues in the study of organizational culture
and performance.
CONCLUSION
This article has shown that the cultural and
behavioral characteristics of organizations
have a measurable effect on a company’s performance.
Organizations with a participative
culture not only perform better than those
without such a culture, but the margin of
difference that widens over time suggests
a possible cause-and-effect relationship between
culture and performance. Much more
research will be needed to substantiate these
findings, but the results are very encouraging.
This research allowed for comparison
between organizations on their cultural
characteristics, management practices, and
key precepts. Clearly, more research needs to
be done along these lines. The numerous and
exciting ideas generated by the recent literature
on organizational culture need to be
tested, so that the importance of this aspect
of organization and management will become
even more evident.
This research also gives a clear vision
of the potential that now exists for
Exhibit 5
DESCRIPTION OF INDUSTRY DATA USED FOR STANDARDIZATION
COMPUSTA T
Industry Industry
Number
of Firms
in Industry
Number
of Firms
in Sample
1ooo Metal mining 20
2200 Textile mill products 44
2600 Paper and allied products 27
1750 Commercial printing 9
2800 Chemicals and allied products 19
2830 Drugs 26
1841 Soap and other detergents 10
2844 Perfumes, cosmetics, toiletries 16
2850 Paints, varnishes, lacquers 9
2911 Petroleum refining 46
3140 Footwear, except rubber 13#p#分頁標題#e#
3221 Glass containers 7
3350 Rolling and drawing non-ferrous metals 16
3531 Construction machinery and equipment 9
3693 X-ray, electromedical apparatus 3
3711 Motor vehicles and car bodies 9
3714 Motor vehicle parts, accessories 27
3720 Aircraft and parts 6
3760 Guided missiles and space vehicles 2
3940 Toys, amusement, sporting goods 15
4210 Trucking - local and long distance 22
4811 Telephone communication 14
4911 Electric services 64
6025 National banks-federal reserve system 90
6798 Real estate investment trust 30
2
1
2
1
2
1
1
1
1
2
1
1
2
1
1
2
2
1
1
1
1
2
1
2
1
monitoring and assessing cultural and behavioral
aspects of organizations. Organizational
diagnoses and cultural audits, for
example, can give an accurate picture of an
organization’s current management system,
and techniques such as human resources accounting
or human-asset accounting can
show the relationship between these characteristics
and the short- and long-term performance
of organizations.
SELECTED BIBLIOGRAPHY
Three readings that emphasize the centrality of zation are Mircea Eliade’s Cosmos and History:
symbolic meaning to the structure of social organi- The Myth of the Eternal Return (Harper & Row, 21
22
1959); George Herbert Meads Mind, Self and
Society (University of Chicago Press, 1934); and
Peter Berger and Thomas Luckman’s The Social
Construction of Reality (Doubleday and Company,
1967). Although these three readings do not
specifically address the concept of organizational
culture, they give an important background for
studying culture and symbolic meaning in organizations.
The importance of values and beliefs in
the organizational literature has long had a normative,
humanistic flavor that is best represented
by the classics: Elton Mayo’s The Human Problems
of an industrial Civilization (Macmillan,
1933); Douglas McGregor’s The Human Side of
Management (McGraw-Hill, 1960); Rensis Likert’s
New Patterns of Management (McGraw-Hill,
1961); and Chris Argyris’s Integrating the Individual
and the Organization (John Wiley and Sons,
1964). Many of the ideas now being developed in
the literature on organizational culture first
appeared in these books.
Corporate culture has recently established
a place in the popular literature that none
of these classics attained. William Ouchi’s Theory
Z (Addison-Wesley, 1981); Thomas J. Peters and
Robert H. Waterman’s In Search of Excellence
(Harper & Row, 1982); Richard Pascale and Anthony
Athos’s The Art of Japanese Management#p#分頁標題#e#
(Simon and Schuster, 1981); Terrence Deal and
Alan Kennedy’s Corporate Cultures (Addison-
Wesley, 1982); and Rosabeth Kanter’s The Change
Musterr (Simon & Schuster, 1983) all brought
the topic to public attention and made an important
contribution to the academic world as well.
Organizational Dynamics (Autumn
1983) and the Administrative Science Quarterly
(September 1983) have each published special issues
on organizational culture. Several other articles,
especially Alan Wilkins and William Ouchi’s
“Efficient Cultures: Exploring the Relationship Between
Culture and Organizational Performance”
(Administrative Science Quarterly, September
1983); Daniel Carroll’s review, “A Disappointing
Search for Excellence” (Harvard Business Review,
November 1983); and Wickham Skinner’s “Big Hat,
No Cattle: Managing Human Resources” (Harvard
Business Review, September 1981) also help to set
a context in which this article may be viewed.
The Survey of Organizations was first
developed in the middle-to-late 1960s. Since that
time it has been revised and updated by David
Bowers and his associates. The best description of
the survey can be found in James Taylor and
David Bowers’ The Survey of Organizations (Institute
for Social Research, 1972).
The conceptualization of organizational
effectiveness used in this article drew on Stanley
Seashore’s article titled, “A Framework for
an Integrated Model of Organizational Effectiveness:’
which appears in Kim Cameron and
David Whetten’s Organizational Effectiveness: A
Comparison of Multiple Models (Academic Press,
1983); Stanley Seashore and Ephraim Yuchtman’s
“Factorial Analysis of Organizational Performance”
(Administrative Science Quarterly, December
1967); Daniel Katz and Robert Kahn’s The Social
Psychology of Organizations (John Wiley and
Sons, 1978); and Gerald H. B. Ross’s unpublished
manuscript “Constituents, Environments and
Organizational Effectiveness” (Institute for Social
Research, 1980).
Finally, the research presented in this article
was drawn from Daniel Denison’s dissertation,
The Climate, Culture and Effectiveness of
Work Organizations: A Study of Organizational
Behavior and Financial Performance (University
of Michigan, 1982).
ACKNOWLEDGMENTS
David Bowers created the data base used for this
study and has sponsored much of my research on
http://www.400190.com/Thesis_Writing/this topic. Stanley Seashore made many useful
comments on early drafts of this manuscript.
Stewart Friedman, Stuart Hart, Doug Cowherd,#p#分頁標題#e#
and Laura Pyle have also made important contributions
to this paper.