The International Journal of Human Resource Management
The International Journal of Human Resource Management
The International Journal of Human Resource Management
Leadership style,
organizational culture and
performance: empirical
evidence from UK
companies
a a
Emmanuel Ogbonna & Lloyd C. Harris
a
Cardiff Business School, University of
Wales, Colum Drive, Cardiff, CF1 3EU, UK
Version of record first published: 18 Feb
2011.
To cite this article: Emmanuel Ogbonna & Lloyd C. Harris (2000): Leadership
style, organizational culture and performance: empirical evidence from UK
companies, The International Journal of Human Resource Management, 11:4,
766-788
This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden.
Abstract The topics of leadership and organizational culture have attracted consider-
Downloaded by [Cornell University] at 12:58 05 November 2012
able interest from both academics and practitioners. Much of the interest in the two areas
is based on explicit and implicit claims that both leadership and culture are linked to
organizational performance. However, while the links between leadership and perform-
ance and between culture and performance have been examined independently, few
studies have investigated the association between the three concepts. This paper examines
the nature of this relationship and presents empirical evidence which suggests that the
relationship between leadership style and performance is mediated by the form of
organizational culture that is present. The paper concludes with a number of implications
for theory and practice.
Keywords Organizational culture; leadership style; performance; United Kingdom.
Introduction
An examination of the literature in the elds of organizational culture and leadership
nds that the two areas have been independently linked to organizational performance.
For example, researchers have examined the links between leadership styles and per-
formance (see Bycio et al., 1995; Howell and Avolio, 1993), and also between
organizational culture and performance (see Deal and Kennedy, 1982; Denison, 1990;
Ouchi, 1981; Pascale and Athos, 1981; Peters and Waterman, 1982; Kotter and Heskett,
1992). Furthermore, numerous aspects of the organizational culture literature allude to
the role of leaders in ‘creating’ and ‘maintaining’ particular types of culture (for
example, Schein, 1992; Siehl, 1985). Equally, the literature on leadership suggests that
the ability to understand and work within a culture is a prerequisite to leadership
effectiveness (see Hennessey, 1998).
However, despite the implicit and explicit linking of leadership and culture in many
parts of organization theory, little critical research attention has been devoted to
understanding the links between the two concepts and the impact that such an
association might have on organizational performance. The absence of critical literature
exploring the performance implications of the links between organizational culture and
leadership is surprising given the numerous references to the importance of the two
concepts in the functioning of organizations (see, for example, Fiedler, 1996; Schein,
1992). The aim of this paper is to provide empirical evidence of the links between
different types of organizational culture, a range of leadership styles and organizational
Literature review
The literature review presented in this paper is examined in three stages. First, studies
Downloaded by [Cornell University] at 12:58 05 November 2012
of the links between leadership style and performance are discussed. Second, research
into the organizational culture–performance link is examined and, nally, studies
combining the analysis of both organizational culture and leadership style are
presented.
An overview of the history of research into the topic of leadership nds that the
literature on leadership and performance can be broadly categorized into a number of
important phases. Early studies on leadership (frequently categorized as ‘trait’ studies
on leadership) concentrated on identifying the personality traits which characterized
successful leaders (Argyris, 1955; Mahoney et al., 1960). Trait theories assume that
successful leaders are ‘born’ and that they have certain innate qualities which
distinguish them from non-leaders (see Stodgill, 1948). However, the dif culty in
categorizing and validating these characteristics led to widespread criticism of this trait
approach, signalling the emergence of ‘style’ and ‘behavioural’ approaches to
leadership (Stodgill, 1948). Style and behavioural theorists shifted the emphasis away
from the characteristics of the leader to the behaviour and style the leader adopted
(Hemphill and Coons, 1957; Likert, 1961). The principal conclusion of these studies
appears to be that leaders who adopt democratic or participative styles are more
successful (see, for example, Bowsers and Seashore, 1966). In this sense, these early
studies are focused on identifying the ‘one best way of leading’.
Similarly to trait theories, the major weakness of style and behavioural theories is
that they ignore the important role which situational factors play in determining the
effectiveness of individual leaders (Mullins, 1999). It is this limitation that gives rise to
the ‘situational’ and ‘contingency’ theories of leadership (for example, Fiedler, 1967;
House, 1971; Vroom and Yetton, 1974) which shift the emphasis away from ‘the one
best way to lead’ to context-sensitive leadership. Although each study emphasizes the
importance of different factors, the general tenet of the situational and contingency
perspectives is that leadership effectiveness is dependent on the leader’s diagnosis and
understanding of situational factors, followed by the adoption of the appropriate style to
deal with each circumstance.
However, in an apparent return to the ‘one best way of leadership’, recent studies on
leadership have contrasted ‘transactional’ leadership with ‘transformational’ leadership.
Transactional leaders are said to be ‘instrumental’ and frequently focus on exchange
768 The International Journal of Human Resource Management
relationship with their subordinates (Bass and Avolio, 1993). In contrast, transforma-
tional leaders are argued to be visionary and enthusiastic, with an inherent ability to
motivate subordinates (Bycio et al., 1995; Howell and Avolio, 1993).
Although the brief summary above indicates that research into leadership has gone
through periods of scepticism, recent interest has focused on the importance of the
leadership role to the success of organizations. Fiedler (1996), one of the most respected
researchers on leadership, has provided a recent treatise on the importance of leadership
by arguing that the effectiveness of a leader is a major determinant of the success or
failure of a group, organization, or even an entire country. Indeed, it has been argued
that one way in which organizations have sought to cope with the increasing volatility
and turbulence of the external environment is by training and developing leaders and
equipping them with the skills to cope (Darcy and Kleiner, 1991; Hennessey, 1998;
Saari et al., 1988). These claims are based on the assumption of a direct link between
leadership and organizational performance. This assumption requires critical review.
Downloaded by [Cornell University] at 12:58 05 November 2012
The earlier review of the literature on the relationship between leadership and
Downloaded by [Cornell University] at 12:58 05 November 2012
performance and between culture and performance nds that many commentators note
that the performance of an organization is dependent on the conscious alignment of
employee values with the espoused values of company strategy. This clearly indicates
that organizational culture and leadership are linked. The following is a review of the
literature on this issue.
One way of uncovering the relationship between culture and leadership is to examine
how culture has been conceptualized in organizational theory. Smircich (1983)
identi es two approaches to the study of the cultural phenomenon in organizations:
culture as an organizational variable, then culture seen as something which can be
manipulated. Thus the nature, direction, and impact of such manipulation are dependent
on the skills and abilities of the leader. The majority of the literature which extols the
virtues of transformational leadership demonstrates widespread support for this view
(for example, Nicholls, 1988; Quick, 1992; Simms, 1997). In contrast, if culture is seen
as an integral part of the organization, then the thinking, feeling, and responses of
leaders are moulded by the culture (Bass and Avolio, 1993; Schein, 1992).
Schein (1992) observes that organizational culture and leadership are intertwined. He
illustrates this inter-connection by looking at the relationship between leadership and
culture in the context of the organizational life cycle. Thus, during the process of
organizational formation, the founder of a company creates an organization which
re ects their values and beliefs. In this sense, the founder creates and shapes the cultural
traits of their organization. However, as the organization develops and time passes, the
created culture of the organization exerts an in uence on the leader and shapes
the actions and style of the leader. Through this dynamic ongoing process, the leader
creates and is in turn shaped by the organizational culture. In summarizing the
consensus of opinion on the links between organizational culture and leadership, Bass
and Avolio (1993) mirror the argument of Schein (1992) by suggesting that the
relationship between the two concepts represents an ongoing interplay in which the
leader shapes the culture and is in turn shaped by the resulting culture.
Bass (1985) demonstrates the relationship between leadership and culture by
examining the impact of different styles of leadership on culture. He argues that
transactional leaders tend to operate within the con nes and limits of the existing
culture, while transformational leaders frequently work towards changing the organiza-
tional culture in line with their vision. Similarly, Brown (1992) observes that good
leaders need to develop the skills that enable them to alter aspects of their culture in
order to improve their organizational performance.
Ogbonna and Harris: Leadership style, organizational culture and performance 771
While there is no shortage of claims that leadership and culture are linked in the
literature (Bass and Avolio, 1993; Nicholls, 1988; Quick, 1992; Schein, 1992; Simms,
1997), there have been very few empirical examinations of the nature and performance
implications of this link. One exception is a recent study of organizational change in the
United States federal civil service. Hennessey concludes that leadership played a major
role in nurturing the appropriate organizational culture which helped to improve the
implementation of speci c government reforms. Hennessey further argues that ‘the
most effective leaders foster, support, and sustain organizational cultures that facilitate
the type of management reform envisioned by “reinventing government” and the
attendant increases in effectiveness and ef ciency’ (1998: 523).
The above review nds that the link between leadership and organizational
performance, the relationship between organizational culture and performance, and the
interplay between leadership and culture have each been studied separately. Inter-
estingly, few empirical studies have combined the simultaneous examination of
Downloaded by [Cornell University] at 12:58 05 November 2012
organizational culture, leadership style, and performance. While some writers suggest
that (1) the style of a leader affects performance, (2) certain types of culture are linked
to superior performance, and (3) culture and leadership are related, the precise nature
and form of interaction between these three concepts is not fully understood. Clearly
further research is necessary to identify, explore, and elucidate the character and pattern
of association between organizational culture, leadership style, and performance.
However, some literature-based conclusions can be drawn. First, the purported
relationship between leadership style and performance is based largely on anecdotal
evidence (Nicholls, 1988; Quick, 1992; Simms, 1997), while the links between
organizational culture and performance are supported by empirical studies (for example,
Gordon and DiTomaso, 1992; Denison, 1990). On the basis of studies which suggest
that leadership style shapes the nature of organizational culture (see, for instance, Bass
and Avolio, 1993; Schein, 1992), it is possible to propose that:
P1 The link between leadership style and organizational performance is mediated
by the nature and form of organizational culture.
In order to guide later discussions this proposition is presented in diagram form in
Figure 1.
Figure 1 The links between leadership style, organizational culture and organizational
performance
772 The International Journal of Human Resource Management
Research design and methodology
To evaluate the ef cacy of the model presented in Figure 1, a descriptive quantitative
research design is clearly appropriate. Consequently, a multi-industry sample of one
thousand units was drawn from the FAME database of registered United Kingdom
rms. Suitable medium- and large-sized rms were selected via the utilization of a
systematic random selection procedure with appropriate units selected on a variety of
criteria, including company turnover, date of registration, and number of employees.
In order to limit potential measurement error, responses were required from key
informants knowledgeable in a variety of tactical and strategic activities (Bowman and
Ambrosini, 1997; Snow and Hrebiniak, 1980; Hambrick, 1981; Nayyar, 1992). While
some researchers argue that the use of a single respondent may be unreliable (see, for
example, Bowman and Ambrosini, 1997), other authors suggest that this issue may not
be a problem in certain contexts (see Zahra and Covin, 1993). Similarly, other
Downloaded by [Cornell University] at 12:58 05 November 2012
researchers have noted the potential negative effect of multiple respondents on usable
response rates (see, for example, Malhotra, 1993), the dif culties of survey administra-
tion (Slater, 1995), and the problems arising from poor inter-rater reliability (see, for
example, Dholakia et al., 1993; Gundlach and Cadotte, 1994). Consequently, it was
decided to adopt a single-respondent approach and a senior executive was selected as a
key informant in each sampling unit.
The ef cient and effective administration and implementation of a survey sig-
ni cantly in uence the overall success of data generation and the achievement of
satisfactory responses (see, for example, Dillman, 1978; Churchill, 1991; Faria and
Dickinson, 1992). Indeed, a plethora of prescriptive articles and books offer helpful
suggestions on effective survey design and execution, including advice on pre-
noti cation protocol, response incentives, and follow-up mailings (see, for instance,
Diamantopoulos et al., 1991; Duncan, 1979; Murphy et al., 1990; Paxson, 1992). In an
effort to improve content validity and response rates the survey was designed,
formulated, and implemented in a manner which closely followed the recommendations
of a variety of authors. In particular, the recommendations on questionnaire design and
layout, survey piloting, and pre-noti cation and post-survey follow-up reminders by
Dillman (1978), Churchill (1991), and Conant et al., (1990) were adopted.
Clearly, a crucial aspect of survey methodology is the development of a questionnaire
which is appropriate for its purpose (Sheatsley, 1983; Churchill, 1991). Consequently,
a research instrument was developed which follows the framework recommended by
Churchill (1991), which in turn is based on the earlier work of Kornhauser and
Sheatsley (1976). The nine-step iterative series of guidelines developed by Churchill
(1991) provides a rigorous systematic procedure for the formulation of a questionnaire
which greatly improves content validity. A review of existing theories, operationaliza-
tions, and measures suggested that the measurement of organizational culture,
leadership style, and organizational performance could be reliably achieved via the
adoption of measures adapted from extant literature.
A variety of organizational culture theorists have presented a number of measures of
organizational culture (see, for example, Cooke and Rousseau, 1988; Xenikou and
Furnham, 1996). A review of these measures indicates that each of the developed
measures of culture re ect the creators’ view and de nition of organizational culture.
Thus, where a theorist de nes organizational culture as a series of values, the measure
of culture focuses on values, whereas de nitions of culture which focus on artefacts
lead to measures which focus on organizational creations (see Harris and Ogbonna,
1999). However, pre-survey discussions with executives revealed that the Deshpande
Ogbonna and Harris: Leadership style, organizational culture and performance 773
et al. (1993) view of culture was consistent with the opinions of practitioners while
concordant with contemporary theory. The measure of organizational culture adapted
from that of Deshpande et al. (1993), which in turn was largely based on the earlier
work of Campbell and Freeman (1991) and Quinn (1988), is viewed as theoretically
superior to other measures on the basis that the battery concentrates on the measurement
of two key continua of organizational culture: organic-mechanistic and internal
maintenance-external positioning. Furthermore, the measure is succinct, easy to
administer, and respondent friendly. Consequently, for practical and for conceptual
reasons the Deshpande et al. (1993) battery was adopted and used in the survey
instrument. However, the labels of culture used by Deshpande et al. (1993) were altered
to the labels competitive, innovative, bureaucratic, and community culture since these
were considered more accurate by practitioners and theoretically justi able by the
researchers. The exact wording of questions and items gauging organizational culture is
presented in Table 1.
Downloaded by [Cornell University] at 12:58 05 November 2012
The measure of perceived leadership style was also derived from extant literature. A
review of literature pertaining to the measurement of leadership behaviour, styles, and
characteristics suggested that a large number of measures might possibly be appropriate
(for example, Fiedler, 1967; Bowsers and Seashore, 1966). However, the measure of
perceived leadership style of House (1971a,b) and House and Dessler (1974), which in
turn was principally based on the earlier work of Fleishman (1957) and Stogdill (1963),
was presented as reliable and valid by a number of respected authors and texts (for
example, Teas, 1981; Kohli, 1989). Indeed, this measure of leadership style has been
widely used in a variety of literatures and is generally accepted as a good measure of
perceptions of leadership style (see Teas, 1981; Kohli, 1989). The precise wording of
questions and items measuring leadership style is presented in Table 2.
It is accepted that business performance is a multi-dimensional and highly complex
phenomenon (Lenz, 1981; Venkatraman and Ramanujam, 1987). While a number of
studies have measured business performance as uni- or bi-dimensional, following the
suggestions of Day and Wensley (1988) and Day and Nedungadi (1994) it was resolved
to gauge performance on dimensions which re ected a broad balance between
customer-focus and competitor-centred perspectives. Consequently, a measure of
organization performance was synthesized from a range of studies adapted from the
constructs of the preceding authors. Performance was measured by analysing long- and
short-term performance constellated around two generic questions complemented by
ve variables which referred to performance in customer satisfaction, sales growth,
market share, competitive advantage, and sales volume.
While a number of past studies had utilized ve-point scales to gauge perceived
leadership style (Kohli, 1989), Barnes et al. (1994) argue that a switch to seven-point
scales has no effect on principal components analysis but often improves the reliability
of answers. Consequently, it was decided to adopt the commonly used seven-point
Likert-type scoring (Likert, 1932a, 1932b) for all items, since the use of seven-point
scales may improve reliability and validity (Churchill and Peter, 1984) as well as all
response rates (Malhotra, 1993).
After two follow-up reminders, 342 responses were received. Unfortunately, twenty
of these responses were ineligible for a variety of reasons, including company
liquidation and inadequate completion of the survey instrument. Following the
recommendations of the Council of American Survey Research Organizations
(CASRO, 1982), response rates were calculated in a manner which removed ineligible
responses from the sample size. This calculation resulted in a return rate of 34.22 per
cent. The majority of respondents were male (87 per cent), the average age was 41.3
774 The International Journal of Human Resource Management
years, the average length of service was nearly six years, while the average length of
time in the current position was nearly four and one half years. No major differences
were found between early and late respondents or between population and sample
industrial classi cations, suggesting that response bias may not be a problem. The
following section details the results of a range of analyses of the responses to the
questionnaire.
Findings
leadership style captured differing dimensions of culture and style. The principal
components analysis of items pertaining to organizational culture (see Table 1) and
items relating to leadership style (see Table 2) were conducted individually. In
accordance with the Kaiser criterion (Kaiser, 1958), factor solutions were retained only
if they exhibited an eigenvalue greater than one and if they were conceptually clear and
interpretable (Churchill, 1991; Hair et al., 1998). It was not necessary to delete items
from the analysis due to lack of variation or because of problems of interpretation. The
identi cation and labelling of the seven extracted factors is discussed below.
Table 1 presents the principal components analysis of measures of organizational
culture adapted from the work of Deshpande et al. (1993), Campbell and Freeman
(1991) and Quinn (1988). As expected, the factor analysis of these items leads to the
extraction of four factors which cumulatively explain nearly 60 per cent of the variance.
The rst factor loads very heavily onto a vector generating an eigenvalue of over three
and accounting for over 22 per cent of the variance. Given that these items appear to
gauge the extent to which an organizational culture is innovative, the solution is
accepted and ascribed the label innovative culture. The second factor solution loads four
items onto a vector generating an eigenvalue of over two and accounting for nearly 16
per cent of total variance. Each of the four items appears to gauge the degree to which
an organizational culture is competitive in nature. Consequently the solution is accepted
and the factor labelled as competitive culture. The penultimate factor solution loads four
items onto a vector generating an eigenvalue of above two. The four items within this
factor solution seem to gauge the degree to which culture is bureaucratic in nature,
leading to the labelling of the factor as bureaucratic culture. The nal factor loads onto
a vector generating an eigenvalue of over one. The four items focus on the extent which
the culture of an organization is focused on the generation and maintenance of an
internal community. The factor is approved and given the label community culture. It
should be noted that the labels adopted for the study correspond (respectively) to the
market, adhocracy, clan, and hierarchy cultures of Deshpande et al. (1993). However,
following discussions with informed practitioners and researchers, the labels com-
petitive, innovative, bureaucratic, and community culture are used, since they are
considered more practically applicable and conceptually valid. These labels are similar
or consistent with the suggestions of a number of organizational theorists (see, for
example, Ouchi, 1980; Mintzberg, 1979; Campbell and Freeman, 1991; Quinn,
1984).
Table 2 presents the principal components analysis of the adapted items of the House
(1971a) and House and Dessler (1974) measure of leadership style. As expected, this
Downloaded by [Cornell University] at 12:58 05 November 2012
solution is accepted and given the label supportive leadership style, with a shorthand
label of supportive leadership. The nal factor solution comprises those items geared
towards the measurement of leadership instrumentality. This measure of leadership
style is akin to directive or transactional leadership and is designed to measure the
extent to which leaders specify expectations, establish procedures, and allocate tasks.
The solution is, therefore, given the label instrumental leadership style, with the
shorthand of instrumental leadership.
Given the widespread use of judgemental and comparative measures of performance,
an index of performance was constructed by calculating the summated mean of all items
for each case. This derived scale was labelled organizational performance. Similarly,
indices were constructed for the four types of organizational culture and the three forms
of leadership style through calculating the mean summated score for all items for each
factor.
Prior to exploring and describing the associations between leadership style, culture,
and performance, it was deemed necessary to gauge the extent of reliability and validity
for each of the indices used in later analysis. Reliability was judged via the calculation
of a Cronbach alpha coef cient (Cronbach, 1951). The calculation of Cronbach alpha
coef cients resulted in alpha coef cients which ranged from 0.6688 to 0.9279 (see
Table 3). Six of the eight measures were above the Nunnally (1978) criterion of 0.7 and
could therefore be classi ed as acceptably reliable without further discussion. However,
two of the measures fall just below this criterion, causing concern. In contrast to the
Nunnally (1978) criterion of 0.7, Peterson (1994) and Slater (1995) suggest that 0.6 is
the ‘criterion-in-use’. This suggestion, coupled with the nding that the deletion of
additional items would merely reduce the coef cient, led to the conclusion that scales
were well above the ‘criterion-in-use’ and thus acceptably reliable.
In addition to the previously discussed efforts to improve content validity during
questionnaire design, formulation and implementation (see Dillman, 1978; Churchill,
1991), to gauge the validation of index operationalization for each measure of items in
the scale were correlated to the whole scale. This analysis indicated signi cant bivariate
relationships in the anticipated direction, indicating convergent validity (see Table 3).
To gauge discriminant validity, an approach akin to that of Gaski (1986) was adopted.
This analysis involves correlating all the measures adopted in the study and gauging the
correlation coef cients between measures against the alpha coef cients of factors (see
Table 4). In the case of both organizational culture and leadership style, no correlation
coef cient is higher than the alpha coef cient of the scale, leading to the suggestion that
the measures adopted in the study exhibit discrimination. Overall, tests of reliability and
Downloaded by [Cornell University] at 12:58 05 November 2012
Factor loading†
Item* Participative Supportive Instrumental Communality
leadership leadership leadership
Before making decisions, s/he considers what her/his subordinates have to say. 0.86698 0.76974
Before taking action s/he consults with subordinates. 0.85945 0.80183
When faced with a problem, s/he consults with subordinates. 0.84408 0.76046
S/he asks subordinates for their suggestions. 0.82952 0.76434
S/he listens to subordinate’s advice on which assignments should be made. 0.81643 0.73284
S/he helps people to make working on their tasks more pleasant. 0.78876 0.71243
S/he looks out for the personal welfare of group members. 0.30527 0.73875 0.64240
S/he does little things to make things pleasant. 0.69752 0.49712
S/he treats all group members as equals. 0.62806 0.57087
S/he explains the way tasks should be carried out. 0.78005 0.61568
S/he decides what and how things shall be done. 2 0.30433 0.69205 0.60737
S/he maintains de nite standards of performance. 0.66436 0.60645
S/he schedules the work to be done. 0.66361 0.52182
Eigenvalues 5.19226 2.00325 1.40784
% Variance explained 39.9 15.4 10.8
Cumulative % variance 39.9 55.4 66.2
Notes
†
Principal components analysis with varimax rotation, converging in seven iterations (all loadings less than 0.3 suppressed).
a
Question wording was ‘Please indicate the extent to which the following statements are true of the Chief Executive Of cer (or equivalent) of your company
by circling the appropriate point’ measured on a 7-point Likert-type scale respectively anchored by (1) Strongly Agree and (7) Strongly Disagree and (1) Very
True and (7) Very False.
Ogbonna and Harris: Leadership style, organizational culture and performance
777
778 The International Journal of Human Resource Management
Table 3 Reliability and scale validation test results for scales of organizational culture and
leadership
Scale Number of Cronbach Inter-item correlations*
scale items alpha coef cient Lowest Highest
Competitive culture 4 0.7692 0.4566 0.6677
Innovative culture 4 0.7642 0.4795 0.6560
Bureaucratic culture 4 0.7022 0.4244 0.6698
Community culture 4 0.6696 0.5233 0.6998
Participative leadership 5 0.9279 0.7936 0.8409
Supportive leadership 4 0.7693 0.4233 0.6751
Instrumental leadership 4 0.6688 0.4222 0.5267
Performance 10 0.8980 0.4161 0.7436
Note
Downloaded by [Cornell University] at 12:58 05 November 2012
* Pearson correlation coef cient. All correlations signi cant at the 0.001 level.
validity lead to the suggestion that the measures adopted and used in later statistical
analyses fall within acceptable reliability and validity criteria.
The initial exploration of data was undertaken by the examination of the descriptive
statistics of measures of culture, leadership style, and organizational performance. As
stated earlier, all items were measured on seven-point scales resulting in a mid-point of
four. The four measures of organizational culture are each somewhat above the mid-
point of four, with reasonable dispersions of central tendency. In contrast, measures of
leadership are somewhat below the mid-point, with the mean for participative
leadership noticeably low. Since it is accepted that objective measures of performance
are highly related to the utilized measures of performance, the comparative high
performance mean may indicate an above-average SBU performance in this study (see
Dess and Robinson, 1984; Pearce et al., 1987; Venkatraman and Ramanujam, 1987).
Proposition 1 suggests that the association between leadership style and performance
is mediated by organizational culture (see Figure 1). In order to explore potential
Figure 2 Path analysis of the links between leadership style, organizational culture and
organizational performance
Ogbonna and Harris: Leadership style, organizational culture and performance 781
The second way in which the effects of the independent factors affect performance is
in the form of effect. The two culture factors exerting a direct in uence and the indirect
effect of the measure of community culture are all positively linked to performance,
whereas the indirect effect of the measure of bureaucratic culture is negative (total
effect 5 0.08). Similarly, two of three measures of leadership style exert an indirect and
positive effect on organizational performance (participative style, total effect 5 0.11;
supportive leadership, total effect 5 0.09). However, the indirect effect of instrumental
leadership is negative (total effect 5 0.08).
Finally, the effects of the independent factors on organizational performance differ in
terms of the impact of total effects. A review of Table 7 demonstrates that the positive
and direct effects of innovative and competitive cultures clearly exert the greatest
impact. In contrast, the indirect association of bureaucratic and community culture is
signi cant but noticeably weaker. Similarly, the effect of the three forms of leadership
style is similar, with a pervasive effect upon performance. These points are important
Downloaded by [Cornell University] at 12:58 05 November 2012
role for senior executives (see Harris and Ogbonna, 1999). Therefore, the nding of
positive associations between externally oriented cultures and performance suggests
that organizational culture change efforts should focus as much on generating external
focus as upon creating internal cohesion and consistency.
The above ndings have implications for the generation of competitive advantage.
Instead of presenting an organization with an advantage over competitors, an internally
oriented organizational culture may prove comparatively disadvantageous when
compared to the advantages possible with externally oriented cultures. Indeed, the
negative links between bureaucratic culture and performance suggest that bureau-
cratization reduces short-term pro tability, impedes long-term growth and may even
affect the survival of the organization. In direct contrast to theories which merely
espouse the generation of widely shared values (for example, Denison 1990; Kotter and
Heskett, 1992) the results of the current study clearly demonstrate that strongly held
values are appropriate only if the culture is geared towards the external environment.
Thus, competitive and innovative cultures which are sensitive to external conditions
have a strong and positive impact on organizational performance. This point provides
some support for the work of Barney (1986, 1991) who argues that for organizational
culture to provide a source of sustainable competitive advantage, the culture must be
adaptable to external contingencies.
As predicted, the associations between the leadership styles studied and organiza-
tional performance are all mediated by some form of organizational culture (see Figure
2). An imprudent interpretation of this nding could be that leadership styles are not
important in relation to performance. This interpretation would clearly be unwise.
Whereas leadership styles are not a strong direct predictor of organizational perform-
ance, the total effects analysis presented in Table 7 indicates signi cant, indirect
pervasive effects on organizational performance. Furthermore, the multiple regression
analyses documented in Table 6 indicate that leadership styles are strong predictors of
both competitive and innovative cultures (which in turn are strong predictors of
performance).
Interestingly, while all of the leadership styles analysed are signi cantly indirectly
associated with performance, instrumental leadership is negatively linked while
supportive and participative leadership styles are positively related (see Table 7). The
nding of a negative indirect link between instrumental leadership and performance
contributes empirical evidence in support of a large number of anecdotal studies which
claim that such a ‘transactional’ leadership style is not consistent with superior
performance (see, for example, Bycio et al., 1995; Bass and Avolio, 1993). The results
Ogbonna and Harris: Leadership style, organizational culture and performance 783
of the current study also contribute to existing knowledge of the effects of supportive
and participative leadership styles. While anecdotal evidence of a direct performance
link was not supported, both supportive and participative leadership styles are positively
associated with innovative and competitive forms of culture. These results indicate that
the generation of an organizational culture, which is externally oriented, is signi cantly
in uenced by the extent to which a leader is supportive of followers and includes
followers in decision-making processes.
A practical implication arising from this study pertains to what has become known as
the ‘managing culture debate’ (Ogbonna, 1993). As is indicated in the literature review,
considerable debate has occurred on the issue of whether organizational culture can be
managed. The overall conclusion of this debate appears to be that the management of
organizational culture is not possible (see Legge, 1994), although certain contingencies
(such as crises or leadership turnover) may present the opportunity to in uence
organizational culture (Martin and Meyerson, 1988). The ndings of this study lead to
Downloaded by [Cornell University] at 12:58 05 November 2012
the suggestion that a potential solution to the dif culties associated with changing
organizational culture may involve focusing on leadership style. While, conceptually
and practically, managing culture is at best dif cult (Ogbonna and Harris, 1998b) and
at worst impossible (Ogbonna, 1993), changes to leadership styles are comparatively
easily achieved. Consequently, while the literature is replete with ‘quick- x’ culture
change programmes designed to improve organizational performance (see Deal and
Kennedy, 1982), the ndings of this paper clearly suggest that, consistent with Fiedler
(1996), leadership-change programmes would be more appropriate.
While this study is designed to provide empirical evidence of the links between
organizational culture, leadership style, and organizational performance, as with most
social science studies, the results of the study raise additional questions, while the
limitations of research suggest alternative avenues for research. Although the cross-
sectional nature of the study precludes claims of causality, additional research adopting
a longitudinal design would provide interesting insight into the direction of associa-
tions. Similarly, while the ndings, implications, and conclusions of this study are
bounded by the context of the research, potentially fruitful research could involve the
replication of this study in a number of different contexts (including speci c industries
or different countries). Finally, the results of this study are limited and constrained by
the measures adopted to gauge organizational culture and leadership style. While the
measures used are accepted as reliable and valid and their selection is defendable,
additional insights into association may be gained by adopting measures of culture and
leadership style which re ect different perspectives. Indeed, while it is hoped that this
study will trigger further debate, without additional research it seems likely that the
links between leadership style, organizational culture, and performance will remain
confusing to both practitioners and theorists. Indeed, Schein notes that ‘leadership and
culture are so central to understanding organizations and making them effective that we
cannot afford to be complacent about either one’ (1985: 327).
References
Porter, L. and McKibben, L.M. (1988) Management and Development: Drift or Thrust into the
21st Century? New York: McGraw-Hill.
Prahalad, C.K. and Bettis, R.A. (1986) ‘The Dominant Logic: A New Linkage Between Diversity
and Performance’, Strategic Management Journal, 7: 485–501.
Quick, J.C. (1992) ‘Crafting an Organizational Culture: Herb’s Hand at Southwest Airlines’,
Organizational Dynamics, 21(2): 45–56.
Quinn, R.E. (1984) ‘Applying the Competing Values Approach to Leadership ’. In Hunt, J.G.,
Stewart, R., Schriesheim, C.A. and Hosking, D. (eds) Managerial Work and Leadership :
International Perspective s. New York: Pergamon Press.
Quinn, R.E. (1988) Beyond Rational Management. San Francisco: Jossey-Bass .
Ray, C.A. (1986) ‘Corporate Culture: The Last Frontier of Control’, Journal of Management
Studies, 23(3): 251–97.
Reed, R. and DeFillippi, R.J. (1990) ‘Causal Ambiguity, Barriers to Imitation, and Sustainable
Competitive Advantage’, Academy of Management Review, 15: 88–102.
Riley, P. (1983) ‘A Structurationist Account of Political Culture’, Administrative Science
Quarterly, 28: 414–37.
Saari, L., Johnson, T.R., McLaughlin, S.D. and Zimmerly, D.M. (1988) ‘A Survey of
Management Education Practices in the U.S. Companies’, Personnel Psychology, 41:
731–43.
Sathe, V. (1983) ‘Implications of Corporate Culture: A Manager’s Guide to Action’, Organiza-
tional Dynamics, Autumn: 5–23.
Schall, M. (1983) ‘A Communication-rules Approach to Organizational Culture’, Administrative
Science Quarterly, 28: 557–81.
Schein, E.H. (1985) Organizational Culture and Leadership, 1st edn. San Francisco: Jossey-
Bass.
Schein, E.H. (1986) ‘What You Need to Know About Your Organizational Culture’, Training and
Development Journal, 40(1): 30–3.
Schein, E.H. (1992) Organizational Culture and Leadership, 2nd edn. San Francisco: Jossey-
Bass.
Scholz, C. (1987). ‘Corporate Culture and Strategy – The Problem of Strategic Fit’, Long Range
Planning, 25(Winter): 3–16.
Sheatsley, P.B. (1983) ‘Questionnaire Construction and Item Writing’. In Rossi, P.H., Wright,
J.D. and Andersen, A.B. (eds) Handbook of Survey Reseach. New York: Academic Press,
pp. 195–230.
Siehl, C. (1985) ‘After the Founder: An Opportunity to Manage Culture’. In Frost, P.J. et al. (eds)
Organizational Culture. Beverly Hills, CA: Sage, pp. 125–40.
Simms, J. (1997) ‘Beauty Queen’, Marketing Business, March: 48–51.
Slater, S. (1995). ‘Issues in Conducting Marketing Strategy Research’, Journal of Strategic
Marketing, 3(4): 257–70.
788 The International Journal of Human Resource Management
Slater, S.F. and Narver, J.C. (1994) ‘Does Competitive Environment Moderate the Market
Orientation –Performance Relationship? ’, Journal of Marketing, 58(January): 46–55.
Smircich, L. (1983) ‘Concepts of Culture and Organizational Analysis’, Administrative Science
Quarterly, 28(September): 339–58.
Snow, C.C. and Hrebiniak, L.G. (1980) ‘Strategy, Distinctive Competence and Organizationa l
Performance’, Administrative Science Quarterly, 25: 317–35.
Stogdill, R.M. (1948) ‘Personal Factors Associated with Leadership’, Journal of Psychology, 25:
35–71.
Stogdill, R.M. (1963) Manual for Leadership Description Questionnaire Form XII, Columbus,
OH: The Ohio State University Bureau of Business Research.
Szymanski, D.M., Bharadwaj, S.G. and Varadarajan, P.R. (1993) ‘Standardization versus
Adaptation of International Marketing Strategy: An Empirical Investigation ’, Journal of
Marketing, 25: 1–17.
Taf nder, P. (1995) The New Leaders: Achieving Corporate Transformation Through Dynamic
Leadership. London: Kogan Page.
Downloaded by [Cornell University] at 12:58 05 November 2012
Teas, R.K. (1981) ‘An Empirical Test of Salesperson’s Job Expectancy and Instrumentalit y
Perceptions’, Journal of Marketing Research, 18(May): 209–26.
Thorlindsson, T. (1987) The Skipper Effect in the Icelandic Herring Industry. Reykjavik:
University of Iceland.
Venkatraman, N. and Ramanujam, V. (1987) ‘Measurement of Business Economic Performance:
An Examination of Method Congruence’, Journal of Management, 13(1): 109–22.
Vroom, V. and Yetton, P. (1974) Leadership and Decision-Makin g. Pittsburgh, PA: University of
Pittsburgh Press.
Weick, K. (1985) ‘Sources of Order in Under Organized Systems: Themes in Recent
Organizational Theory’. In Lincoln, Y.S. (ed.) Organizational Theory and Inquiry: The
Paradigm Revolution. Beverly Hills, CA: Sage.
Weiner, Y. (1988) ‘Forms of Value Systems: A Focus on Organizational Effectiveness and
Cultural Change and Maintenance ’, Academy of Management Review, 13(4): 534–45.
Willmott, H. (1993) ‘Strength is Ignorance: Slavery is Freedom: Managing Culture in Modern
Organizations ’, Journal of Management Studies, 30(4): 515–51.
Xenikou, A. and Furnham, A. (1996) ‘A Correlational and Factor Analysis Study of Four
Questionnaire Measures of Organizational Culture’, Human Relations, 49(3): 349–71.
Zahra, S.A. and Covin, J.G. (1993) ‘Business Strategy, Technological Policy and Firm
Performance’, Strategic Management Journal, 14: 451–78.