Content uploaded by Torben Pedersen
Author content
All content in this area was uploaded by Torben Pedersen on Jan 28, 2014
Content may be subject to copyright.
Working Paper
The impact of knowledge management on MNC
subsidiary performance:
the role of absorptive capacity
By
Volker Mahnke, Torben Pedersen
and Markus Venzin
No. 1-2004
Institut for Informatik Department of Informatics
Handelshøjskolen Copenhagen
i København Business School
Howitzvej 60 Howitzvej 60
2000 Frederiksberg DK-2000 Frederiksberg
Denmark
Tlf.: 3815 2400 Tel.: +45 3815 2400
Fax: 3815 2401 Fax: +45 3815 2401
http://www.inf.cbs.dk
http://www.inf.cbs.dk
© Volker Mahnke, Torben Pedersen and Markus Venzin
The impact of knowledge management on MNC
subsidiary performance: the role of absorptive
capacity
Volker Mahnke, Torben Pedersen and Markus Venzin
Volker Mahnke, Associate Professor, Dep. of Informatics, Copenhagen Business
School, Howitzveij 60 (4); 2000 Copenhagen-FDB; vm.inf@cbs.dk, Denmark
Torben Pedersen, Professor, Dep. Of International Management and Economics,
Copenhagen Business School, Howitzveij 60 (2); 2000 Copenhagen-FDB;
tp.intbs.dk, Denmark
Markus Venzin, Assistant Professor, Dep. Of Strategy, Bocconi University, Via
Isonzo 23; 20136 Milano;
mvenzin@uni-bocconi.it
, Italy
Abstract
Empirical studies on the impact of knowledge management on the performance of MNC
subsidiaries remain elusive to date. This study examines the effect of knowledge
management tools such as corporate university, communities of practice, group
benchmarking, learning systems and rewards upon absorptive capacity and performance
with unique data from subsidiary units in a large German MNC – Heidelberger Cement.
The findings suggest that knowledge management tools unfold their performance
impact through their significant influence on absorptive capacity and knowledge
inflows. The key contributions to the current literature on knowledge flows in the MNC
include an empirically corroborated link between deployments of knowledge
management tools and their impact on the subsidiary employee’s ability and motivation
to learn from internal knowledge flows in the MNC as well as their impact on
subsidiary business performance.
1
Introduction
This study examined the effect of applying knowledge management tools in subsidiaries on
developing absorptive capacity to benefit from internal knowledge flows to improve business
performance with unique data from a German MNC – Heidelberger Cement. Knowledge
management tools include reward systems, learning system, and communities of practice,
group benchmarking report, and corporate university. Proxies measuring subsidiary
employee’s motivation and ability to learn from knowledge-inflows captured absorptive
capacity. Business performance was assed by several dimensions of business performance
including economic gains and product development. Knowledge management tools strongly
predicted absorptive capacity indicating managerial discretion in the development of a unit’s
absorptive capacity; absorptive capacity as endogenous variable strongly predicts knowledge
inflow and the knowledge inflow predict significantly business performance on a subsidiary
level. Findings suggest integration of research on knowledge-flows, absorptive capacity, as
well as subsidiary performance and a concomitant stimulant for future research along these
lines.
Many authors, picturing the modern MNC as a "differentiated network," have discussed
the importance of knowledge flows between subsidiaries for the MNC performance (Bartlett
and Ghoshal, 1989; Gupta and Govindarajan, 1991; 2000). However, only a few studies have
started to examine this relationship empirically, looking mainly at the effect of antecedence
to knowledge sharing and learning in the external (e.g. Lyles and Salk, 1996; Lane and
Lubatkin, 1998; Lane, Salk and Lyles, 2001, Simonin, 1999) and internal relations of the
MNC (e.g. Gupta and Govindarajan, 2000; Hansen, 1999; Holm and Pedersen 2000;
Minbaeva et al, 2003). Common to these studies is that they focus on antecedence of
knowledge sharing and assess the level of shared knowledge. However, with a few
exceptions like Andersson et al. (2002) the relation between knowledge flows and subsidiary
performance remains largely asserted rather than empirically corroborated. This study
addresses the performance implications of knowledge flows in the MNC directly.
Several studies on the MNC knowledge flows propose that the absorptive capacity of
the receiving unit is the most significant determinant of internal knowledge transfer in the
MNC (e.g. Gupta and Govindarajan, 2000). When subsidiaries differ in their absorptive
2
capacity, this affects the level of knowledge transfer from other MNC units (Cohen and
Levinthal, 1990). While the literature offers multiple methods to conceptualise and
operationalize absorptive capacity (Zahra and George, 2002), little attention has been paid to
the question of whether and how firms can enhance the development of absorptive capacity
through deploying knowledge management tools. With a few exceptions (e.g. Foss and
Pedersen, 2002; Minbaeva et al, 2003) empirical studies treat absorptive capacity as given
and as an exogenous determinant to knowledge processes. Little is known about how
knowledge management tools influence absorptive capacity and how this impacts knowledge
inflows and subsidiary performance.
Most empirical studies of knowledge processes on subsidiary level are based on surveys
with only one informant (typically the CEO or another top-manager) from each
organization/subsidiary. The implicit assumption behind this methodology is that perceptions
on knowledge processes are so homogeneous inside the organization that one central
informant can express “the average” perception of all members of the organization. This is a
highly questionable assumption and in this paper is instead applied a unique data-set
collected on the level of individuals that includes individual perceptions of the use of
knowledge management tools, motivation, ability, knowledge inflows and subsidiary
performance.
The contribution of this article is three-fold: 1) It introduces a model that treats
subsidiary absorptive capacity as an endogenous construct in enabling knowledge inflows
and enhancing performance on a subsidiary level. Of course, the message that absorptive
capacity matters for knowledge inflows is hardly new (e.g., Cohen and Levinthal, 1990), but
systematic theorizing and testing of knowledge tools as enablers of absorptive capacity are
elusive; 2) While most studies have restricted their analyses to effects on knowledge flows to
the neglect of its performance implications this article includes a direct test of performance
implications of knowledge inflow on subsidiary level; and 3) On the methodology, a unique
data-set including individual perceptions of knowledge processes from one single MNC –
HeidelbergCement – is applied to test the whole model including the managerial actions
(Knowledge Management tools), the intermediary steps (developing absorptive capacity and
increasing knowledge inflows) and the final outcome (subsidiary performance).
3
Theory and hypothesis
Below, we pursue two arguments. The first argument relates to the conceptualisation and
operationalisation of absorptive capacity on a MNC subsidiary level. We explain why
absorptive capacity matters for intra-firm knowledge flows, and how levels of absorptive
capacity can be influenced by the deployment of knowledge management tools. In this
argument we picture absorptive capacity as the motivation and ability of subsidiary
employees to deal with knowledge inflows, both may be actively influenced by
management’s selection of knowledge management tools. The second argument deals with
the performance implications of knowledge inflows that are enabled through absorptive
capacity. Is there any reason to expect that levels of knowledge inflows will vary in their
effect on subsidiary performance? The following section develops our arguments to derive
hypotheses as summarized in the following model.
Figure 1. The theoretical model
Rewards
Learning-
system
Communities
of practice
Benchmarkig
Corporate
university
Absorptive capacity:
- ability and motivation
to learn
H2
H1 H3
Knowledge
Inflow
Business
Performance
Rewards
Learning-
system
Communities
of practice
Benchmarkig
Corporate
university
Absorptive capacity:
- ability and motivation
to learn
H2
H1 H3
Knowledge
Inflow
Business
Performance
Developing subsidiary absorptive capacity
The concept ‘absorptive capacity’ has been mainly used to capture a company’s ability
to recognize, assimilate, and apply external knowledge to commercial ends (Cohen and
Levinthal, 1990: 128). Extending the work of Lane and Lubatkin (1998), Lane, Salk and
4
Lyles (2001) in their study on knowledge acquisition in international JV propose that “the
first two components, the ability to understand external knowledge and the ability to
assimilate it, are interdependent yet distinct from the third component, the ability to apply the
knowledge” (p.1156).
Zahra and George (2002) criticised existing studies for applying measures (like R&D
intensity, number of scientists working in R&D departments, etc.) that “have been
rudimentary and do not fully reflect the richness of the construct” (p. 199). For example,
current measures of absorptive capacity are focused on the absorption of knowledge in the
MNC’s external relations rather than on the absorption of knowledge from other subsidiaries
in the MNC; they neglect the role of individuals in the organization, but their motivation and
ability is crucial for knowledge utilization and exploitation. Furthermore, current measures
may be too occupied with the ability to recognize and assimilate external knowledge (e.g.
potential absorptive capacity) but neglect the role of the receiving unit’s motivation to put
knowledge to commercial use (e.g. realized absorptive capacity).
Addressing these critiques, Minbaeva et al (2003) identify the subsidiary’s employees’
ability and motivation as the key aspects of a subsidiary’s absorptive capacity that in turn
enables a subsidiary to benefit from internal knowledge flows. Clearly, when employees of a
subsidiary lack the ability to understand and recognize knowledge of other units, for example
because they are cognitively focussed in a specialized knowledge domain (March, 1991;
Levinthal and March, 1993) a subsidiary’s absorptive capacity will decline. Likewise, when
employee’s lack motivation to deal with new knowledge, --- for example due to the not-
invented here syndrome, or because they face arcane, complicated, and time consuming
social interaction with other units, --- the result may be knowledge-hostility between
subsidiaries that negatively affects a subsidiary’s absorptive capacity (Michailova and
Husted, 2003). Conversely, if subsidiary employee’s are both able and motivated with
regards to acquiring external knowledge this indicates higher absorptive capacity and in
accordance with Cohen and Levinthal, (1990) and Zahra and George (2002) we expect:
H1: Higher absorptive capacity increases knowledge inflows
While many empirical studies regard absorptive capacity as an exogenous construct to
knowledge inflows, this study regards it as an endogenous construct subject to degrees of
5
managerial discretion (see also Foss and Pedersen, 2002, Minbaeva et al, 2003). Here we
consider different knowledge management tools, which may contribute to the development of
the absorptive capacity. This allows us to examine the possible managerial influence on
absorptive capacity that is not often examined in the literature. How do knowledge
management tools employed at Heidelberger Cement including learning system, and
communities of practice, corporate university, subsidiary benchmarking, and reward systems
influence the subsidiary’s employees’ ability and motivation to utilise knowledge flows from
other units in the MNC?
As suggested by much research on knowledge management (see Argote, 1999; Krogh
et al, 2000 for a synthesis) learning systems need to provide access to other MNC unit’s
knowledge. Only when MNC employee’s can identify valuable knowledge by searching
topics or knowledgeable partner’s in other MNC units without incurring substantial
transaction costs (Mahnke and Venzin, 2003; Kautz and Mahnke, 2003) will they be able and
motivated to utilize knowledge inflows. By contrast, when individual transaction costs of
knowledge search and access are substantial, for example because adequate communication
channels are missing (Pedersen et al, 2003), individual’s motivation will decrease
accordingly. To decrease transaction cost of knowledge inflows Heidelberger Cement
deployed an intranet-solution which acts as facilitator for identifying experts and strategically
important knowledge domains. In addition, expert teams identify leading practices in diverse
knowledge areas such as energy efficiency, new concrete applications, and logistics to
develop benchmarking reports that codify and document valuable knowledge created in
leading subsidiaries so that other subsidiary’s can improve their performance based on
benchmarking.
While explicit knowledge sharing is enabled through benchmarking reports, the
corporate university, and learning system, knowledge sharing in communities and teams
complements such efforts, but tends to focus more on tacit knowledge exchange (Leonard-
Barton and Sensiper, 1998). Several forms of teamwork, including communities of practice
are conducive to the integration and creation of knowledge in Heidelberger Cement (Wenger
and Snyder, 2000). Interdisciplinary teams often integrate knowledge (Grant, 1996) that
hitherto existed separately and dispersed across function in the MNC. Through interactive
learning, community members develop community specific communication codes or
combinative capabilities (Kogut & Zander, 1993; Monteverde, 1995). Thus, by educating
6
employees across subsidiary boundaries “communities of knowing” emerge, and through
specific language codes their members increase their ability and motivation (Osterloh and
Frey, 2000) to combine and blend a variety of knowledge across the MNC’s subsidiaries.
Teamwork in communities does not only educate a subsidiary’s employees, it also
enhances their involvement and motivations to better utilise knowledge inflow’s from other
community members, which are employed at other subsidiaries. Through integrating
knowledge of individual community members, teams may not only blend knowledge and
insights beyond what individual members may achieve; new knowledge development may
also be stimulated by conversations and language based learning in teams (e.g. Brown &
Duguid, 1991; Boland & Tenkasi, 1995). Thus, teamwork also fuels knowledge creation
resulting in “new combinations” across subsidiaries (Schumpeter 1934; Galunic and Rodan,
1998). In sum then, the analysis of communities of practice suggests that they influence a
subsidiary’s absorptive capacity through increasing the motivation and ability of subsidiary’s
employee’s to share knowledge in social interaction.
Opinions on the influence of reward systems (both formal and informal) on the ability
and motivation of a subsidiary’s employees with regards to knowledge sharing remain
divided. On the one hand, several authors argue the motivation for exchanging knowledge
between subsidiaries cannot be easily influenced through explicit rewards (Osterloh and Frey,
2000; Krogh et al, 2000; Wenger and Snyder, 2000; Bartlett and Ghoshal, 1989; Ouchi,
1982). First, for explicit rewards to be effective motivators requires a sound measurement
base (Ouchi, 1982). However, where input, output, or processes cannot be standardized and
by implication easily measured, explicit incentives rewarding knowledge sharing behaviour
of subsidiary employee’s become easily arbitrary. This may be the case where knowledge-
sharing processes are surrounded by substantial causal ambiguity and uncertainty (Simonin,
1999). If so, providing explicit rewards on an arbitrary measurement base can lead to the
perception of unfair incentives to crowd out intrinsic motivation of subsidiaries employee’s
(Osterloh and Frey, 2000).
In addition, when knowledge sharing behaviour is multidimensional in that it requires
initiatives along several dimensions including active requests, networking, and building close
social ties (Hansen, 1999) explicit rewards that focus on one dimension to the neglect of
another may undermine motivation to engage in complementary activities (Holmström and
Milgrom, 1991). By implication, knowledge sharing may rather strive if participants in
7
horizontal knowledge flows between subsidiaries are intrinsically motivated by common
goals (e.g. Ouchi, 1082; Osterloh and Frey, 2000). Thus, signalling (Spence, 1972)
organizational support through informal acknowledgement may help intrinsic motivation
without incurring the negative effects of misdirected extrinsic rewards.
On the other hand, employees take part in knowledge sharing only as long as the
benefits exceed the costs; otherwise, they may withdraw. Accordingly, whenever possible
increasing the employee’s benefits through providing explicit rewards is appropriate. For
example, when subsidiary employees’ knowledge sharing behaviour can be specified in less
uncertain and ambiguous knowledge exchange relation, providing explicit rewards alongside
informal acknowledgements can increase motivation, in particular, if explicit rewards act as
complement rather than as substitute to intrinsic motivation (Frey, 1997; Foss and Mahnke,
2003; Laursen and Mahnke, 2001). In addition, as shown by Mahnke and Venzin (2003)
developing a measurement base for explicit rewards through monitoring by experts can also
be seen as an investment to increase common shared knowledge between subsidiary
employee’s, which in turn increases the ability to share knowledge among them. In sum
then, we expect that knowledge management tools act as antecedence to absorptive capacity
defined as the motivation and ability of employee’s to acquire new knowledge, thus:
H2: Learning system, communities of practice, corporate university, benchmarking, and
reward- systems influence positively the absorptive capacity.
Knowledge inflows and subsidiary performance
A subsidiary’s employees’ may be able and motivated to acquire and assimilate knowledge
from other MNC units, but might not have the capability to transform and exploit the
knowledge to benefit the subsidiary’s bottom line (e.g. Nelson and Winter, 1982; Hedlund,
1994; Szulanski, 2003). As a necessary condition, knowledge inflows must influence the
receiving units employee’s understanding as well as their behaviour and action (Fiol and
Lyles, 1985). Only when knowing-doing gaps are closed can knowledge sharing among
subsidiary’s lead to cost reduction and innovation. For that purpose, the receiving unit often
needs to adapt received knowledge to local circumstance, however (Hedlund, 1994).
8
While the influence on individual’s action is a minimum condition, knowledge sharing
may benefit the receiving units through routine adaptation involving several individuals
(Nelson, and Winter, 1982; Szulanski, 2003). For example, Nelson & Winter (1982: 131)
suggest that ‘reliable routines of well understood scope’ provide the best components for
intra-firm replication of processes, and perhaps more importantly, new combination of
routines for innovations. Nonaka and Takeuchi (1995) agree that knowledge inflow’s among
units involves the social processes of linking new knowledge to existing knowledge (e.g.
about components or processes). In sum then, knowledge inflows may impact a subsidiary’s
performance through changing individual insights and behaviour, be it individually or
embedded in organizational routines (e.g. Kim, 1993).
While knowledge sharing may contribute to value creation in a subsidiary, it has also
its costs. For example, several authors suggest that these costs may be substantial and include
time spent on knowledge sharing and adaptation, infrastructure investments etc (see for a
review, Foss and Mahnke, 2003), as well as opportunity costs in terms of expenditures and
cognitive capacity that could have been used to foster knowledge creation and process
improvement internally to the focal subsidiary (Gammelgaard and Pedersen, 2003). Thus,
being able and motivated to put knowledge from other MNC units to use is one thing. It may
be quite another thing to benefit from knowledge inflows. Nonetheless, we expect:
H3: Knowledge inflows influence positively a subsidiary’s performance
Data and Method
Research Site and Data Collection
The research was conducted in a German cement company, HeidelbergCement. With around
1’500 subsidiaries in over 50 countries, worldwide cement sales for over 6 billion Euro and
36,000 employees, HeidelbergCement is one of the four largest cement manufacturers in the
world. During our study, the company entered in a consolidation phase after rapid expansion.
HeidelbergCement grew from below 10’000 employees to the current size in less than 10
years, mainly through acquisitions. An objective of the gradual integration of the newly
acquired companies was to increase the knowledge exchange between the different units.
9
Although one site-sampling has disadvantages in terms of generalizability of results across
sectors, it has advantages through controlling for contextual factors that complicate the
interpretation of results in multiple-firm samples (i.e. variation in industry and firm-specific
factors). In addition, previous research has successfully used single firm sampling to explore
the effects of intra-firm networks on resource exchange and combination (Tsai and Ghoshal,
1998). Thus, we decided to limit our survey to one company, which allowed us to include
more informants from each subsidiary. This is an advantage compared to most other studies
that only include one informant from each organization and treat the perception of this
informant as representative for the whole organization (e.g. Lane and Lubatkin, 1998; Gupta
and Govindarajan, 2000).
We developed a questionnaire based on a literature review and following our first
phase of 52 on-site interviews with senior mangers from the 6 business units of
HeidelbergCement. The interviews ranged from 1 to 4 hours in length. We sent the survey to
10 subsidiary managers as a pilot run to increase the clarity of our questions and avoid
interpretation errors. We discussed the face validity of the questions, which resulted in
several substantial changes of the phrasing of the questions. After the survey pre-testing, we
identified a sample of 265 leading MNC unit managers stemming from all geographical
business areas of HeidelbergCement. The selected managers represented 3 distinct groups:
the first group included 100 managers that had either been previously interviewed by our
research team or participated in one of the international knowledge management initiative
fostered by the MNC’s headquarter. The second group consisted of 100 mangers that have
not been interviewed and did not participate in knowledge management workshops. In
addition, we included 65 engineers from the subsidiaries to make sure to get the views of the
employees that are most directly involved in international day-to-day knowledge transfer.
The questionnaire was sent out to participants in paper form together with a 6-page
brochure explaining the scope of the knowledge management initiatives at HeidelbergCement
as well as a letter by the CEO supporting the survey. Participants had 4 weeks to return the
questionnaire via mail or fax. This mailing process was managed by the headquarter’s office.
We collected 222 usable questionnaires – a response rate of 83,77%. The fact that the
participants had to indicate their names on the envelop of the survey most likely contributed
to this high return rate.
10
Measures
All data were collected through the questionnaire and most variables are multi-item
measures that were measured using seven-point Likert scales. However, items such as the
number of employees were measured using actual values. The following sections provide the
exact wording used for questionnaire items.
Business performance. Several researchers have turned to perceptual measures of
performance as better forward-looking measure of performance than more objective
performance measures (e.g.) that capture past performance rather than future performance
(e.g. Andersson et al., 2001). It has also been shown that perceptual measures tend to have a
high correlation with objective accounting-based measures (Geringer and Hebert, 1991).
Accordingly, we apply a perceptual measure of performance and our definition of business
performance captures many dimensions of business performance including economic gains
and product development. The respondents were asked to what extent the applied knowledge
management initiatives in HeidelbergCement were expected to improve business results in
terms of cost saving, additional revenue, process improvements, product improvements and
new product and processes. Respondents indicated this on a seven-point Likert scale, where 1
was defined as “not important” and 7 was defined as “very important” for all the five above-
mentioned activities. Business Performance is a multi-item construct calculated as the
average score reported by respondents across these five items (Alpha=0.87).
Inflow of knowledge. The construct of “Inflow of knowledge” captures the extent of
inflow of knowledge from other units in HeidelbergCement to the focal unit. The construct
was measured by asking respondents to assess to what extent does knowledge represented in
other unit’s influences your work, where 1 equaled “not at all” and 7 equaled “very strongly.
Absorptive Capacity. The construct of “absorptive capacity” captures both the
individual perception of ability to learn and acquire new knowledge, and the motivation to do
so. The ability to learn was measured by asking respondents to assess the following two
statements “I perfectly understand the knowledge” and “I can easily acquire new
knowledge”. In order to measure the motivation to learn the respondents were asked to assess
the following two statements “I find knowledge sharing rewarding” and “I don’t fear they
will steal my ideas” (reverse coded). All four statements were measured on a seven-point
11
Likert scale, where 1 was defined as “never” and 7 was defined as “always”. Taken together
these four items measures the Absorptive Capacity as a multi-item construct. The construct
was calculated as the average score reported by respondents across these four items
(Alpha=0.83).
Reward. This construct is capturing the perception of the individuals as to what extent
the organization (HeidelbergCement) is rewarding individual knowledge activities. The
perception of the rewarding system was measured by asking respondents to assess the
following two statements “My efforts to create and share knowledge are rewarded formally”
and “My efforts to create and share knowledge are rewarded informally”. Both were
measured on a seven-point Likert scale (1 equaled “don’t agree” to 7 equaled “strongly
agree”) and the construct of rewarding was calculated as the average score of these two items
(Alpha=0.63).
Learning culture. The construct of “learning culture” captures the individual
perception of the openness towards knowledge sharing and the availability of needed
knowledge in the organization. In order to measure the learning culture the respondents were
asked to assess the following three statements “I know whom to ask when I need
knowledge”, “I know where to look when I need knowledge” and “I have easy access to
communication channels” on a seven-point Likert scale (1 equaled “don’t agree” to 7 equaled
“strongly agree”). The construct of learning culture was calculated as the average score of
these three items (Alpha=0.71).
Corporate university. The construct of “corporate university” is capturing the
individual perception of the importance of a knowledge management tool introduced by
HeidelbergCement, namely a corporate university. The respondents were asked to assess the
importance in terms of impact on: 1) personal field of activities, 2) business results, 3)
knowledge sharing, and 4) knowledge creation. Our measure of the significance of Corporate
university is based on the average of these four items (Alpha=0.82).
Group benchmark report. The construct of “group benchmark report” captures the
individual perception of the importance of a knowledge management tool introduced by
HeidelbergCement, namely group benchmark report. The respondents were asked to assess
the importance in terms of impact on: 1) personal field of activities, 2) business results, 3)
12
knowledge sharing, and 4) knowledge creation. Our measure of the significance of Group
benchmark report is based on the average of these four items (Alpha=0.71).
Communities of practice. The construct of “communities of practice” captures the
individual perception of the importance of a knowledge management tool introduced by
HeidelbergCement, namely communities of practice. The respondents were asked to assess
the importance in terms of impact on: 1) personal field of activities, 2) business results, 3)
knowledge sharing, and 4) knowledge creation. Our measure of the significance of
Communities of practice is based on the average of these four items (Alpha=0.76).
Controls. To control for the characteristics of knowledge, particularly the codification
vs. tacitness of knowledge, that have been identified as an important factor influencing the
inflow of knowledge in other studies (e.g. Kogut and Zander, 1992; Szulanski, 1996;
Simonin, 1999) we have measured the extent to which the organizational knowledge is
represented in a codified vs. tacit form. The level of codification of organizational knowledge
is measured by asking respondents to assess the following statements for knowledge
represented in their unit “New employees can easily learn what we know”, “Knowledge is
extensively documented and well structured” and “Knowledge in our unit can be easily
expressed”. The three items are added together and the average of the three items are making
up our construct, level of codification (Alpha=0.68).
Furthermore, two more control variables are related to the business performance. These
are the extent to which the knowledge in the focal unit are formally acknowledged by other
organizational units in HeidelbergCement (measured as a single item construct on a seven-
point Likert scale where 1 equaled “not at all” and 7 equaled “to a high degree”) and the
extent to which knowledge from other units must be adapted in order to apply it in the
context of the focal unit (measured as a single item construct on a seven-point Likert scale
where 1 equaled “not at all” and 7 equaled “to a high degree”). We expect that formal
acknowledgement of knowledge in the focal unit will have a positive impact on business
performance, while required knowledge adaptation will have a negative impact on business
performance.
13
Results
Tests of Hypotheses
The hypotheses may be summarized in three basic models as follows.
1) Absorptive capacity = Reward + Learning culture + Communities of practice + Group
benchmark report + Corporate university + Error
2) Inflow of knowledge = Absorptive capacity + Error
3) Business performance = Inflow of knowledge + Error
However, since the above models represent decisions that are interdependent (i.e., they have
to be considered jointly), the use of single equation models may yield biased results and
obscure interesting theoretical possibilities. As the above models are interdependent, it is
possible that the joint optimization of all involved variables may lead to sub-optimization of
one or more individual variables. Statistically, the interdependence might be reflected in that
error terms of the three models are somehow correlated. Hence, the correct model to estimate
these decisions is a simultaneous equation model as three-stage least square, which
circumvents the problem of interdependence by using instrument variables (often the
exogenous variables) to obtain predicted values of the endogenous variables (in our case:
knowledge inflow, and business performance).
The correlation matrix (including all correlation coefficients) and descriptive data
(mean values, standard deviation, minimum and maximum values) on all variables in the
model are provided in the Appendix 1. In order to detect potential problems of
multicollinearity we should look at correlation coefficients among variables 1-5 (1. equation),
variables 6-7 (2. equation) and variables 8-10 (3. equation). None of these are above the usual
threshold indicated the possibility of multicollinearity (i.e. r>0.5), (Hair et al., 1995).
However, the correlations between corporate university, communities of practice and group
benchmark report gets close with coefficients 0.41-0.49. But, running the models with some
of the correlated variables omitted had no effect on the explanatory power of the main
variables. Therefore, we concluded that the results are very stable in terms of the different
14
specifications of the model and the data doesn’t seem to include problems of
multicollenearity.
We have applied the three-stage least square regression techniques (3SLS) with
instrument variables to test all hypotheses simultaneously. The result of the total model is
reported in Table 1 and figure 2. Numbers in parentheses represent standard errors.
Table 1: The three-stage least squares estimation of a simultaneous equation model.
Equations
ABSORPTIVE
CAPACITY
INFLOW OF
KNOWLEDGE
BUSINESS
PERFORMANCE
Intercept 0.77*
(0.45)
1.81
(1.17)
0.11
(0.86)
Reward 0.03
(0.03)
Learning culture 0.09*
(0.05)
Communities of practice 0.17***
(0.06)
Group benchmark report 0.11***
(0.04)
Corporate university 0.09**
(0.04)
Absorptive capacity
1.10***
(0.34)
Codification of knowledge -0.07
(0.05)
Inflow of knowledge 0.94***
(0.19)
Formal acknowledgement
0.01
(0.05)
Knowledge adaptation
0.06
(0.04)
F-value
R-square
N
4.18***
222
8.73***
222
6.85***
0.25
222
***, ** and * = significant at 1, 5 and 10 percent, respectively.
Overall, the system of the three equations (models) works well, with a system weighted
R-square of 0.25. This indicates that almost one fourth of the observed variation in the
business performance is explained by the variables in the model. We turn now to the tests of
our explanatory hypotheses.
15
As shown in Table 1 three knowledge management tools of group benchmark report (p
<1%), communities of practice (p < 1%) and corporate university (p < 5%) have a
significantly positive impact on the development of absorptive capacity in the focal unit. The
same is true for the learning culture (p < 10%) that is positively facilitating the absorptive
capacity. However, the reward turns out to be insignificant (but with the expected positive
sign). The results indicate that there is managerial discretion in developing the absorptive
capacity, however, some knowledge management tools and incentive structures will have a
stronger effect than others. A more explicit reward system (extrinsic motivation) seems to be
of less value in the development of absorptive capacity, while learning culture and the
knowledge management tools like group benchmark report and corporate university seems to
have a strong positive impact. This is in line with the expanding literature on intrinsic vs.
extrinsic motivation in knowledge management (e.g. Osterloh and Frey, 2000).
Furthermore, as expected absorptive capacity have a strong positive impact (p < 1%) on
the inflow of knowledge from other units and this is true even when controlling for the level
of codification (turned out to be insignificant in itself). And the inflow of knowledge has a
very strong positive impact (p < 1%) on business performance, while both the two control
variables turned out to become insignificant.
Figure 2. The empirical model with significant coefficients
16
These findings can be juxtaposed against two prominent strands of research. Empirical
research on knowledge flows has investigated impediments and facilitators of knowledge
flows in firms, but is interested mainly in whether and to what degrees knowledge has been
transferred given particular constellations of impediments and facilitating factor. Knowledge
transfer, however, is costly and knowledge management tools deployed may be seen as an
investment to increase business performance on several dimension. Thus, to examine the
This study produced major evidence for the contention that deployment of knowledge
management tools has important implications for the performance of MNC subsidiaries.
Applying theory on knowledge flows and conceptualizing absorptive capacity as the ability
and motivation to learn from new knowledge on the subsidiaries level, we found that
knowledge management tools impact a subsidiary’s absorptive capacity. Importantly, results
indicated that absorptive capacity is subject to managerial discretion because knowledge-
management tools, such as group benchmarking, communities of practice, learning culture
etc, significantly influence levels of absorptive capacity. We showed also, that absorptive
capacity influences knowledge inflows, which significantly influence subsidiary
performance, thus eliminating doubts about their causal impact. This study showed that the
application of knowledge management tools could be an important source of competitive
advantage in the MNC, especially when they increase the ability and motivation of the
subsidiary employees to utilize knowledge inflows from other units.
The advantage of the model is that it establishes a link between the individual
knowledge activities and the business performance, where business performance is defined as
a multi-dimensional construct capturing not only the efficiency gains but also production and
product developments. The model shows that there is a strong link between the micro-
processes of setting up incentive structures and applying knowledge management tools
through managerial action on the level of the unit and the business performance. The link
goes through the development of absorptive capacity and the stimulation of knowledge
inflow from other units.
Discussion and Conclusion
17
Second, future research could also distinguish between intra-firm and inter-firm sources of
knowledge flows to assess their impact on business performance. For instance, the
performance impact of knowledge flows might differ if sources of knowledge are external
rather than internal to the MNC. The current study did focus on internal sources of
knowledge flows only but neglected external knowledge inflows that figure so importantly in
MNC research (e.g. Cantwell and Mudambi, 2003).
Weaknesses in the present study provide some suggestions for future research. First,
prospective research could explore whether and how a subsidiary’s motivation and ability to
learn does influence a subsidiary’s knowledge sharing behavior. Often subsidiaries receive
knowledge inflows based on their own knowledge sharing behavior with other units (Gupta
and Govindarajan, 1991, 2000). An important question of future research could be to
investigate knowledge inflows and outflows of focal MNC units simultaneously.
The present study has obvious affinity to recent research on absorptive capacity. In
general our findings support the assertion that absorptive capacity facilitates knowledge
inflows. In contrast to prior empirical studies that seek to investigate knowledge absorption
between firms, the current study was interested in intra-firm knowledge flows between MNC
subsidiaries. Accordingly, following Minbaeva et al (2003) we conceptualized a subsidiary’s
absorptive capacity and developed measures to capture the ability and motivation of
employees to learn from other units in the MNC. Clearly, absorptive capacity defined and
measured this way opens new avenues for further research on how antecedence influence
absorptive capacity, and how it impacts knowledge flows in both internal and external
relations of the MNC.
performance impact of knowledge management it is not enough to examine whether
knowledge has been transferred. The current study is, to the best of our knowledge, the first
to investigate not only antecedence of, but also the impact of knowledge flows on subsidiary
performance. Clearly, researchers should no longer treat performance of knowledge
management as an assumption and black box. Rather, opening the black box to unpack the
relation between knowledge management tools employed, the knowledge flows they enable,
and their impact on business performance does constitute an important area of future
research.
18
Lastly, future research could empirically uncover the importance of alternative control
mechanisms on the impact of knowledge inflows and outflows on a subsidiary level. The
present study dealt with selected knowledge management tool that influenced the focal
subsidiaries absorptive capacity. While our findings indicate managerial discretion in
building absorptive capacity through knowledge management tool, somewhat surprisingly,
we have not found strong support on the impact of reward systems on absorptive capacity. By
exploring different types of extrinsic and intrinsic motivators (e.g. Osterloh and Frey, 2000),
future research could investigate under what conditions these influence positively the ability
and motivation of subsidiary employees to absorb and effectively utilize knowledge flows
from internal and external sources to increase performance.
19
20
References
Andersson, U., M. Forsgren & U. Holm. 2002. The strategic impact of external networks:
Subsidiary performance and competence development in the multinational
corporation. Strategic Management Journal, 23 (11): 979-998
Andersson, U., M. Forsgren & T. Pedersen. 2001. Subsidiary performance in multinational
corporations: the importance of technology embeddedness. International Business
Review, 10: 3-23.
Argote, L. 1999. Organizational learning: Creating, retaining and transferring knowledge.
Norwell, MA: Kluwer.
Bartlett, Christopher & Sumantra Ghoshal. 1989. Managing Across Borders. Boston:
Harvard Business School Press.
Boland, R. J. & Tenkasi, R.V. 1995. Perspective making and perspective taking in
communities of knowing. Organization Science, 6: 350-372.
Brown, J.S. & Duguid, P. 1991. Organizational learning and communities-of-practice:
Towards a unified view of working, learning, and innovation. Organization Science,
Vol. 2, No. 1, February, 40 - 57.
Cantwell, J. & R. Mudambi. 2003. Multinational enterprises and competence-creating
knowledge flows: A theoretical analysis. In: Mahnke & Pedersen, Knowledge Flows,
Governance, and the Multinational Enterprise - Frontiers in international Management
Research. Palgrave Macmillan. Forthcoming.
Cohen, Wesley M. & Daniel A. Levinthal. 1990. Absorptive Capacity: a New Perspective
on Learning and Innovation. Administrative Science Quarterly, 35:128–152
Fiol, Marlene C. & Marjorie A. Lyles. 1985. Organizational Learning. Academy of
Management Review, 10(4): 803-813
Foss, N and V. Mahnke. 2003. Knowledge management: What can organizational economics
contribute? In: Lyles, M and Esterby, M. (2003) Handbook of organizational learning
and knowledge management: 78-103. Blackwell Publisher.
Foss, Nicolai & Torben Pedersen. 2002. Transferring Knowledge in MNCs: The Roles of
Sources of Subsidiary Knowledge and Organizational Context. Journal of
International Management, 8: 1-19.
Frey, B. 1997. Not just for money. A Personal Theory of Motivation. Edward Elgar
Galunic, D. C., & Rodan, S. 1998. Resource re-combinations in the firm: Knowledge
structures and the potential for Schumpeterian innovation. Strategic Management
Journal, 19: 1193-1201.
Gammelgard, J., U. Holm & T. Pedersen.2003. The Dilemmas of MNC Subsidiary Transfer
of Knowledge. In: Mahnke & Pedersen, Knowledge Flows, Governance, and the
Multinational Enterprise - Frontiers in international Management Research. Palgrave
Macmillan. Forthcoming.
Geringer, J. M. & L. Hebert (1991). Measuring performance of international joint ventures.
Journal of International Business Studies, 22(2), 249-263.
Grant, R.M. (1996). Toward a Knowledge-Based Theory of the Firm. Strategic Management
Journal, 17(special issue):109-22.
Gupta, A.K. and Govindarajan, V. (1991). Knowledge flows and the structure of control
within multinational firms. Academy of Management Review, 16(4): 768-792.
Gupta, A.K. and Govindarajan, V. (2000). Knowledge flows within multinational
corporations. Strategic Management Journal, 21(4): 473-496.
21
Hansen, M. T. 1999. The search transfer problem: The role of weak ties in sharing
knowledge across organizational subunits. Administrative Science Quarterly 44: 82-
111.
Hedlund, G. 1994. A model of knowledge management and the n-form corporation. Strategic
Management Journal, 15: 73 - 90.
Holm, Ulf & Torben Pedersen. 2000. The Emergence and Impact of MNC Centres of
Excellence. London: Macmillan Press.
Holmström, B. & Milgrom, P. 1991. Multitask principal-agent analyses: Incentive contracts,
asset ownership, and job design, Journal of Law, Economics, and Organization 7: 24-
52.
Kautz. K. and V. Mahnke. 2003. Value Creation through IT-supported Knowledge
Management? The Utilisation of a Knowledge Management System in a Global
Consulting Company. Informing Science 6: 75-88.
Kim, D. H. 1993. The link between individual and organizational learning. Sloan
Management Review 35, 1: 37-49.
Kogut, B. & Zander, U. 1993. Knowledge of the firm and the evolutionary theory of the
multinational corporation. Journal of International Business Studies, 24: 625-645.
Kogut, B., & Zander, U. 1992. Knowledge of the firm, combinative capabilities, and the
replication of technology. Organization Science, 3: 383-397.
Krogh, G. von, Ochijo, K , Nonaka I., Enabling Knowledge Creation, Oxford University
Press, 2000.
Lane, Peter & Michael Lubatkin. 1998. Relative Absorptive Capacity and Interorganizational
Learning. Strategic Management Journal, 19(5): 461-477
Lane, Peter, Jane E. Salk & Marjorie A. Lyles. 2001. Absorptive Capacity, Learning and
Performance in International Joint Ventures. Strategic Management Journal, 22(12):
1139-1161.
Laursen K. and V. Mahnke, 2001. Knowledge strategies, innovative capacity and
complementarity in human resource practices. Journal of Management and
Governance 1: 1-27.
Leonard-Barton, D. & Sensiper, S. 1998. The role of tacit knowledge in group innovation.
California Management Review 40/3: 112-132.
Levinthal, D. A., & March, J. G. 1993. The myopia of learning. Strategic Management
Journal 14: 95-113.
Lyles, Marjorie A. & Jane E. Salk. 1996. Knowledge Acquisition from Foreign Partners in
International Joint Ventures. Journal of International Business Studies, 27(5): 877-
904
Mahnke, V. and M. Venzin (2003) Governing knowledge teams in the MNC. Management
International Review: forthcoming
March, J. G. 1991. Exploration and exploitation in organizational learning. Organization
Science, 2: 71-87.
Michailova, S., & K. Husted. 2003. Knowledge-sharing hostility in Russian firms. California
Management Review. Berkeley, 45 (3): 59-81.
Minbaeva, D., T. Pedersen, I. Björkman, C. Fey and H.J. Park (2003). MNC Knowledge
Transfer, Subsidiary Absorptive Capacity and HRM. Journal of International
Business Studies, forthcoming.
Monteverde, K. 1995. Technological dialog as incentive for vertical integration in the semi
conductor industry, Management Science, 41: 1624-16-1638.
22
Nelson, R. & Winter, S. 1982. An evolutionary theory of economic change. Cambridge, MA:
The Belknap Press.
Nonaka, I. & H. Takeuchi. 1995. The knowledge creating company. Oxford University Press.
Osterloh. M. and B. Frey, 2000. Motivation, knowledge transfer and organizational forms.
Organization Science, 11(5): 538-550.
Ouchi, W. G. 1982. Markets, Bureaucracies, and Clans Administrative Science Quarterly, 25,
129-141.
Pedersen, T., B. Petersen & Deo Sharma (2003). Knowledge Transfer Performance of
Multinational Companies. Management International Review, forthcoming.
Schumpeter, J.A. 1934. The Theory of Economic Development. Cambridge, MA: Harvard
Press.
Simonin, Bernard L. 1999. Transfer of Marketing Know-How in International Strategic
Alliances: an Empirical Investigation of the Role and Antecedents of Knowledge
Ambiguity. Journal of International Business Studies, 30(3): 463-490
Spence, A. M. 1972. Job market signalling. Quarterly Journal of Economics, 87. 355-77.
Szulanski, Gabriel. 1996. Exploring Internal Stickiness: Impediments to the Transfer of Best
Practice within the Firm. Strategic Management Journal, 17(Winter): 27-43.
Szulanski, G. 2003. Adaptation of know how for cross-border transfer. Management
International Review: forthcoming
Tsai, W. and S. Ghoshal. (1998). Social capital and value creation: The role of intrafirm
networks. Academy of Management Journal, 41(4); 464-477.
Wenger E. and Snyder W. 2000. Communities of Practice: The Organizational Frontier,
Harvard Business Review: 78(1): 139-145.
Zahra, Shaker. A. & Gerard George. 2002. Absorptive Capacity: a review,
reconceptualization, and extension. Academy of Management Review, 27(2): 185-203
Appendix 1. Correlation matrix for all variables included in the model (N=222)
1 2 3 4 5 6 7 8 9 10 11
1) Reward 1.00
2) Learning culture 0.29*** 1.00
3) Communities of practice 0.10 0.15** 1.00
4) Group benchmark report 0.12* 0.17** 0.41*** 1.00
5) Corporate University 0.15** 0.19*** 0.45*** 0.49*** 1.00
6) Absorptive capacity 0.19*** 0.17** 0.22*** 0.12* 0.11 1.00
7) Codification of knowledge -0.28*** -0.27*** -0.10 -0.19*** -0.18** -0.13* 1.00
8) Inflow of knowledge 0.16** 0.26*** 0.23*** 0.16** 0.22*** 0.15** -0.24*** 1.00
9) Formal acknowledgement 0.22*** 0.25*** 0.04 0.13* 0.07 0.06 -0.25*** 0.12* 1.00
10) Knowledge adaptation -0.05 -0.03 0.12* 0.05 0.05 -0.04 -0.15** 0.05 0.08 1.00
11) Business performance 0.06 0.14** 0.47*** 0.60*** 0.56*** 0.10 -0.14** 0.20*** 0.13* 0.13* 1.00
Mean 3.74 5.37 5.29 4.76 5.05 3.23 4.17 5.05 4.32 3.94 5.06
Std. Dev. 1.27 0.84 0.92 1.17 1.07 0.88 1.04 1.20 1.43 1.20 0.78
Min. values 1 2.66 2.25 1.75 1 1 1 2 1 1 2.13
Max values 7 7 7 7 7 7 6.33 7 7 7 7
***, ** and * indicates 1%, 5% and 10% significance level, respectively
