ArticlePDF Available

Deliberate Learning in Corporate Acquisitions: Post-Acquisition Strategies and Integration Capability in US Bank Mergers


Abstract and Figures

This paper introduces a knowledge-based view of corporate acquisitions and tests the post-acquisition consequences on performance of integration decisions and capability-building mechanisms. In our model, the acquiring firm decides both how much to integrate the acquired firm and the extent to which it replaces this firm's top management team. It can also learn to manage the post-acquisition integration process by tacitly accumulating acquisition experience and explicitly codifying it in manuals, systems, and other acquisition-specific tools. Using a sample of 228 acquisitions in the U.S. banking industry, we find that knowledge codification strongly and positively influences acquisition performance, while experience accumulation does not. Furthermore, increasing levels of post-acquisition integration strengthen the positive effect of codification. Finally, the level of integration between the two merged firms significantly enhances performance, while replacing top managers in the acquired firm negatively impacts performance, all else being equal. Implications are drawn for both organizational learning theory and a knowledge-based approach to corporate strategy research. Copyright © 2004 John Wiley & Sons, Ltd.
Content may be subject to copyright.
Strategic Management Journal
Strat. Mgmt. J.,25: 1233– 1256 (2004)
Published online in Wiley InterScience ( DOI: 10.1002/smj.426
INSEAD, Fontainebleau, France
The Wharton School, University of Pennsylvania, Philadelphia, U.S.A.
This paper introduces a knowledge-based view of corporate acquisitions and tests the post-
acquisition consequences on performance of integration decisions and capability-building mech-
anisms. In our model, the acquiring rm decides both how much to integrate the acquired
rm and the extent to which it replaces this rm’s top management team. It can also learn
to manage the post-acquisition integration process by tacitly accumulating acquisition experi-
ence and explicitly codifying it in manuals, systems, and other acquisition-specic tools. Using
a sample of 228 acquisitions in the U.S. banking industry, we nd that knowledge codica-
tion strongly and positively inuences acquisition performance, while experience accumulation
does not. Furthermore, increasing levels of post-acquisition integration strengthen the positive
effect of codication. Finally, the level of integration between the two merged rms signicantly
enhances performance, while replacing top managers in the acquired rm negatively impacts
performance, all else being equal. Implications are drawn for both organizational learning the-
ory and a knowledge-based approach to corporate strategy research. Copyright 2004 John
Wiley & Sons, Ltd.
The performance of corporate acquisitions has long
been a topic of interest to researchers in several
disciplines, such as industrial economics, manage-
ment, and nance. Considerable heterogeneity still
exists, however, with respect to the denition of
performance (e.g., benet to the acquiring rm,
the acquired rm, the combined entity) as well as
to its measurement (accounting returns, stock price
reactions, etc.). Overall, evidence shows that stock-
holders of the acquired rms make positive eco-
nomic returns, while acquirers’ abnormal returns
Keywords: M&A; mergers; post-acquisition integration;
top management; learning; knowledge; codication bank-
Correspondence to: Maurizio Zollo, INSEAD, Department of
Strategy & Management, 77305 Fontainebleau, France.
(in either nancial or accounting terms) are not
statistically distinguishable from zero. Whereas the
evidence on the average magnitude of value cre-
ated for the various counterparts involved is rela-
tively uncontroversial, the explanation of the vari-
ance around the mean is still very much in need
of both theoretical and empirical work.
In this paper, we focus on acquirers’ variation
in performance and examine how learning
processes specic to the management of the
post-acquisition phase affect it. We also provide
a theoretical argument and an empirical test
for the performance implications of post-
acquisition integration decisions, as well as the
interaction between these decisions and some
resource- and capability-based antecedents. Our
focus on the acquiring rm, instead of the
target or the combined entity, is inuenced
by the observation that learning processes and
Copyright 2004 John Wiley & Sons, Ltd. Received 4 October 1999
Final revision received 5 May 2004
1234 M. Zollo and H. Singh
post-acquisition decisions are housed primarily
within the acquirer’s corporate development
department or its relevant business unit.
The U.S. banking industry, where the study is
positioned, is a good example of a particularly tur-
bulent environment, where the tight coupling of
deregulation, disintermediation, and technological
evolution has generated an unprecedented wave of
acquisitions in a relatively short amount of time. It
thus provides a good laboratory for testing whether
different acquirers’ approaches to post-acquisition
management and different levels of expertise in
managing the integration process are systemati-
cally associated with different performance out-
The section below summarizes relevant prior
research. Subsequently, we introduce a knowledge-
based perspective, which we apply to the manage-
ment of acquisition processes, and advance testable
hypotheses. The following section then describes
the research design, some of the key ndings in
our eldwork, and the operationalization of the
most important theoretical constructs. Finally, we
discuss the results of our analyses and conclude
by noting several implications of our ndings for
theories of corporate strategy and organizational
Here, we examine how scholars in nancial eco-
nomics, strategic management, and organizational
theory have discussed and tested the performance
implications of corporate acquisitions.
The market for corporate control and the
resource-based views of acquisitions
Research in nancial economics has examined
returns to acquirers and targets in large samples
of acquisitions. In general, it views acquisitions as
transactions that reect the market for corporate
control, in which management teams vie for the
control of rms’ productive assets. If one man-
agement team underperforms, then a more com-
petent team takes its place (Manne, 1965; Jensen
and Ruback, 1983). Empirically, this research has
found that although there are positive gains from
the combination of the acquiring rm and the tar-
get’s assets, most of these gains accrue to share-
holders of the target rm. More recently, it has
shown that average abnormal returns to the acquir-
ing rm are either statistically equivalent to zero
(Jarrell, Brickley, and Netter, 1988; Franks, Har-
ris, and Titman, 1991; Loderer and Martin, 1992;
Shleifer and Vishny, 1994; Agrawal and Jaffe,
2000) or lower (Agrawal, Jaffe, and Mandelker,
1992). This result is conrmed in the literature
on acquisitions in the banking industry developed
by nancial economics scholars (Hawawini and
Swary, 1990; Rhoades, 1994). Rhoades’ review of
39 bank merger studies, in particular, showed that
changes in acquiring rms’ accounting and nan-
cial returns around the acquisition event are not
statistically signicant.
The strategic management eld, on the other
hand, has focused on several factors that might
inuence the post-acquisition performance of ac-
quirers. Most prominently, it has used the resource-
based view of the rm (Wernerfelt, 1984; Rumelt,
1984; Barney, 1986; Dierickx and Cool, 1989) to
test the impact of resource relatedness on such
performance (Chatterjee, 1986; Singh and Mont-
gomery, 1987; Lubatkin, 1987; Shelton, 1988;
Seth, 1990; Healy, Palepu, and Ruback, 1992;
Chatterjee et al., 1992). The evidence, however,
suggests that no clear relationship links resource
relatedness and performance. This variation in
results may be explained in several ways. First,
there might be several mechanisms that can inu-
ence the post-acquisition performance of the com-
bined entity, but that do not rely on the exploitation
of economies of scale and scope, and that therefore
would not benet from higher degrees of related-
ness between the two rms. Seth (1990), for exam-
ple, found evidence for coinsurance effects, which
allow the combined entity to obtain higher leverage
by combining uncorrelated streams of cash ows
to yield higher tax shields. Additionally, Baker and
Montgomery (1994) observed that LBO rms and
‘enlightened conglomerates’ can consistently cre-
ate signicant rents by developing idiosyncratic
capabilities in the structuring of highly powered
incentive systems (LBO rms) and in restructur-
ing, turnaround, and control processes.1Second,
Barney (1988) argues, consistent with evidence
shown by Lubatkin (1987) and Singh and Mont-
gomery (1987), that an acquirer has to create
1For another example of how acquirers can consistently cre-
ate value without recourse to synergistic potential from resource
relatedness, see the case of Hanson Plc (Taubman and Haspes-
lagh, 1992).
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1235
a uniquely valuable and inimitable combination
of its assets with those of the acquired rm to
earn positive abnormal returns on its investment.
Although many acquirers may possess resources
related to those of a target, the uniqueness con-
dition provides a much more stringent criterion
for value creation. Third, although relatedness may
sometimes directly and signicantly impact per-
formance, as in the context of consolidating and
declining industries (Anand and Singh, 1997),
these conditions might not be generalizable to
other industry conditions. Thus, relatedness may
be a necessary but not sufcient requirement for
superior performance. Generally, it might be pos-
sible to achieve synergy only when rms carefully
design and execute integration processes focused
on extracting the gains associated with the combi-
nation of the two organizations. Accordingly, when
resource relatedness is used to explain how eco-
nomic rents may accrue to acquirers, it is important
to include as explanatory variables the activities
necessary to extract such rents.
This paper intends to improve our understand-
ing of the predictors of acquisition performance by
exploring the explanatory roles of pre-acquisition
resource characteristics, learning from prior acqui-
sitions, and key elements of the post-acquisition
integration process. By design, however, it restricts
the scope of the strategic intent that could poten-
tially be pursued through acquisitive growth by
focusing on horizontal acquisitions among com-
mercial banks, where the economic logic can be
dened in terms of either cost efciencies derived
from larger scale in the same geographic markets,
or by a combination of (more limited) cost ef-
ciencies and revenue enhancement potential from
the expansion of activities in new geographic mar-
kets. Hence, product diversication strategies (e.g.,
the acquisition of an investment bank or insurance
business) and new product development objectives
(e.g., acquisitions of nancial ‘boutiques’ special-
ized in the structuring or trading of innovative
products) are excluded from the scope of the study.
Also, in this study, acquisition performance is
dened as the variation in the acquiring rm’s
overall performance and measured as the deviation
from competitors’ long-term variation in return on
assets.2The choice of the accounting measure to
2Please see the Measures paragraph in the section on Research
Methodology and Measures for details on the measurement of
the dependent variable in this study.
proxy the performance construct is forced by the
relative unobservability of post-acquisition deci-
sions, as well as learning processes, from the nan-
cial market standpoint. Because nancial markets
are unlikely to be able to anticipate and incorporate
information relative to our key explanatory vari-
ables in the acquirer’s stock price at or around the
time of the acquisition announcement, event-study
models that focus on cumulative abnormal returns
to stock prices are not appropriate for our analysis.
The next section reviews the extant literature that
focuses on how rms manage the post-acquisition
Research on the management of integration
The process acquiring rms use to manage their
acquisitions is substantially more complex to study
empirically, compared to the relatedness studies
reviewed above, because of the lack of process-
level data typically available for a sufciently large
number of observations. As a result, prior research
in this area has established few denitive ndings.
Jemison and Sitkin (1986) indicate that it is use-
ful to think about acquisitions in terms of both
their strategic and organizational t, which gener-
ally do not correspond neatly to each other. Thus,
the organizational complexity of an acquisition can
be quite different from the strategic considerations
driving the transaction. Building on this insight,
Haspeslagh and Jemison (1991) highlight the rel-
evance of the processes rms use to select their
acquisition targets, negotiate the agreement to pur-
chase or to merge, decide how to manage the post-
acquisition transition phase, and interact with the
acquired rm to implement the selected integration
strategy. They also indicate some critical dimen-
sions of the post-acquisition decision-making pro-
cess, such as the extent of functional integration
and the timing for its implementation. Their work
was an important step in understanding the dimen-
sions of the integration process and in relating the
strategic objectives driving the acquisition to key
managerial decisions made in the post-acquisition
phase of the transaction.
Subsequent work has attempted to understand
post-acquisition processes by focusing on one deci-
sion at a time. Consequently, it trades off rich
contextual descriptions of the interdependencies
among integration decisions for analytical pre-
cision and theoretical rigor. For example, Pablo
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1236 M. Zollo and H. Singh
(1994) examined the antecedents of the decision
about the level of integration, whereas Datta and
Grant (1990) and Shanley (1994) attempted to test
the performance implications of this decision and
found some support for a positive inuence on per-
formance. All these authors dene the construct
‘level of integration,’ drawing on Thompson’s
(1967) pioneering work, as the extent to which
the functions of the acquired unit are linked to,
aligned with, or centralized in, the equivalent func-
tions of the acquiring organization, and we make
no exception in this study.3More recently, Capron
(1999) focused her attention on a related phe-
nomenon—the extent of resource redeployment
and knowledge transfer between the two organi-
zations—and found that it is signicantly related
to increased performance. This provides additional
evidence that achieving some degree of integration
between the two organizations offers economic
Another important dimension of the post-acqui-
sition integration process involves the degree to
which pre-existing resources within the acquired
rm are replaced with the equivalent resources
of the acquirer, or are simply dismissed. Chief
among these resources is the human and social
capital embedded in the acquired rm’s employ-
ees, particularly in its top management team. Con-
trary to the predictions of the ‘market for cor-
porate control’ perspective, Cannella and Ham-
brick (1993) found that managerial turnover was
harmful to acquisition performance and that the
impact increased in magnitude when more senior
managers were replaced. More recently, Krishnan,
Miller, and Judge (1997) reached similar conclu-
sions, and noted that the degree of complemen-
tarity between the two top management teams
positively inuences performance and should be
protected when possible.
Other studies have researched the antecedents
of the decision to replace the target’s top manage-
ment team. Walsh (1988) examined top manage-
ment turnover rates by comparing post-acquisition
turnover in a sample of rms to that of a con-
trol group. He found that turnover rates could not
be explained by the product market relationship
3In the context observed, acquisitions in the banking industry,
the concept of ‘level of integration’ translates into a set of fairly
discrete choices related to decisions such as the conversion of
information systems, the alignment of loan approval processes,
or the rationalization of the two networks of bank branches.
between the acquirer and the target rm. In subse-
quent work, Walsh and Ellwood (1991) found that
post-acquisition turnover is inuenced by the pre-
acquisition protability of the acquirer, rather than
that of the target, as one would expect. In partic-
ular, the higher was the acquirer’s pre-acquisition
performance, the lower was the post-acquisition
turnover of the acquired company’s management.
In sum, this research has emphasized the poten-
tial benets and the complexities involved in cre-
ating value through acquisition processes. Striking
the right balance between achieving the necessary
level of organizational integration and minimiz-
ing the disruptions to the acquired rm’s resources
and competencies is a fundamental challenge that
affects the success not only of the integration pro-
cess but also of the entire acquisition. Because of
these managerial trade-offs, it is important to better
understand whether and how rms develop pro-
cesses and capabilities specic to the management
of corporate acquisitions.
The empirical work that explicitly considers the
relationship between the acquiring rm’s experi-
ence and acquisition performance shows that sim-
ple ‘learning curve’ explanations are of limited
relevance. Although some studies have found that
such experience positively impacts performance
(Fowler and Schmidt, 1989; Bruton, Oviatt, and
White, 1994), others have found no such rela-
tionship (Lubatkin, 1987; Baum and Ginsberg,
1997). Haleblian and Finkelstein (1999) reported
evidence for a non-linear, U-shaped, relationship,
which highlights possible negative learning effects
(Gick and Holyoak, 1987) for the rst few acqui-
sition experiences, during which acquirers might
inappropriately apply lessons learned in past expe-
riences to contexts that seem supercially similar
but are inherently different, thereby reducing the
probability of success. In a similar vein, Hayward
(2002) nds no linear impacts of prior acquisition
experience on short-term stock price reactions, but
a number of non-linearities in the quality of such
experience (such as the average success of prior
acquisitions). Finally, evidence of non-linear expe-
rience effects in multi-task contexts (i.e., spillovers
of alliance experience on acquisition performance)
was found by Zollo and Reuer (2003).
In the next section, we present a knowledge-
based perspective on the management of acqui-
sition processes and show that the lack of con-
sistency in empirical tests of the learning curve
hypothesis might be due to incomplete theoretical
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1237
treatment of the underlying organizational learning
processes, rather than to anomalies of the M&A
context or broader unobserved heterogeneity.
The knowledge-based view of the rm (Nelson
and Winter, 1982; Kogut and Zander, 1992; Grant,
1996) suggests that the outcome of the acquisi-
tion process is inuenced by the degree to which
the acquiring rm develops a capability specicto
managing the acquisition process. Prior literature
has highlighted this capability as a key prerequi-
site for completing these complex organizational
endeavors successfully (Haspeslagh and Jemison,
1991: Ch. 2). Building on their initial insight, we
intend to rst develop a theoretical understand-
ing of the mechanisms that might underlie this
collective learning process. We then apply these
notions to the development of one specic type
of acquisition-related capability, i.e., managing the
process through which the acquired rm is, par-
tially or totally, integrated within the structures
and processes of the acquiring rm. We note that
the integration capability is not the only one of
relevance to the success of the post-acquisition
phase. The ability to identify the appropriate acqui-
sition candidate, for example, is just as important
as the ability to integrate it once acquired. Whereas
acquirers can develop competence related to differ-
ent aspects of the acquisition process, in this paper
we focus on the integration capability since this is
likely to be a crucial antecedent to the performance
of the acquisition, and has so far not received spe-
cic attention from a theoretical standpoint. We
therefore consider other competencies unrelated to
the integration process outside the scope of the
paper and indicate them as promising areas for
future research.
Organizational knowledge and capability-
building mechanisms
Previous literature on acquisitions has used
research on learning curves originally developed
to understand manufacturing processes (Yelle,
1979; Dutton and Thomas, 1984; Epple, Argote,
and Devadas, 1991; Lapre, Mukherjee, and Van
Wassenhove, 2000) to test whether learning
processes exist within acquirers. It linked the
accumulation of experience in prior acquisition
processes with improvements in acquisition
performance as measured by either nancial
variables or survival (Lubatkin, 1987; Fowler and
Schmidt, 1989; Bruton et al., 1994; Pennings,
Barkema, and Douma, 1994; Baum and Ginsberg,
1997; Haleblian and Finkelstein, 1999; Hayward,
2002). Firms might be able to learn how to
manage acquisition processes by simply doing
more of the same, and thereby tacitly forming and
rening organizational routines that might directly
(i.e., without explicit knowledge articulation or
codication) impact the performance of subsequent
acquisitions. This ‘learning-by-doing’ hypothesis
can be more formally stated and submitted as
Hypothesis 1: The greater the acquiring rm’s
previous acquisition experience, the better the
economic performance of the focal acquisition.
The accumulation of prior experience, however, is
not the only way in which rms can develop col-
lective capabilities in handling organizational tasks
(Zollo, 1998; Kale, Dyer and Singh, 2002; Zollo
and Winter, 2002). We argue that one reason why
the prior literature on learning effects in acquisi-
tions has not derived consistent results is that it has
failed to account for mechanisms different from
‘learning-by-doing’ to explain how rms improve
in their understanding of the ways acquisitions
should be managed. We therefore intend to lever-
age on, and possibly extend, the work produced
by scholars interested in the strategic implications
of organizational knowledge to discuss the role
of more rened mechanisms in explaining orga-
nizational learning processes in contexts such as
corporate acquisitions.
Rogers (1980), Winter (1987), and Kogut and
Zander (1992) propose several dimensions of orga-
nizational knowledge that inuence how practices
evolve and transfer within and across rms. These
dimensions include the degree to which knowl-
edge is articulable, teachable, and codiable, or
the extent to which the individuals and the groups
which possess the knowledge are actually aware
of it, can describe it, and therefore communi-
cate it using oral or written media (Polanyi, 1962,
1966). These dimensions are clearly interrelated.
For example, the degree of articulability and teach-
ability will inuence the degree of codiability.
In the context of acquisition management, it is
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1238 M. Zollo and H. Singh
likely that the knowledge underlying any given
organizational process can accumulate in both
explicit forms, such as manuals, blueprints, infor-
mation systems, and implicit forms, such as human
Given the same level of codiability of knowl-
edge necessary to perform a certain task, how-
ever, rms might choose to codify the amount
of accumulated experience to different degrees.
For instance, rms with equivalent levels of expe-
rience might develop different kinds of written
tools or information systems related to the man-
agement of the acquisition processes. Not all cod-
iable and teachable knowledge is actually codi-
ed and taught. Because the costs of creating and
updating tools and systems are likely to be high,
the proportion of codied knowledge to what is
potentially codiable might be quite small. The
decision to invest scarce managerial resources in
knowledge codication processes might therefore
be interpreted as a strategically relevant activity,
which could signicantly affect the development
of explicit task-related competence.
As a group produces tools and systems to exe-
cute a given task, it will have to evaluate how and
why its past decisions and actions for similar situa-
tions have inuenced performance. This effort will
likely improve the quality of the group’s under-
standing of the causes of successes and failures in
the task at hand. It will, in other words, increase a
rm’s capability to plan and manage that particu-
lar process. Weick’s (1995) work on retrospective
sensemaking evokes well how we believe capabil-
ity building happens in the context we observed.
Nevertheless, we do not assume that the group that
develops and renes these codes learns intention-
ally. Firms, or groups of individuals, might very
well learn about the drivers of performance in their
acquisitions without realizing that they are doing
so. For example, the development of an integra-
tion manual is normally motivated by the need to
coordinate the execution of the huge number of
virtually simultaneous activities necessary to align
processes in the two organizations. In doing so,
however, the acquirer might develop a theory of
what decisions are most appropriate in what condi-
tions, therefore unintentionally contributing to the
improvement of the understanding of the perfor-
mance implications of those decisions.
If these arguments are correct, the degree to
which past experience is reected upon, articulated,
and codied into ad hoc tools should inuence
how effective organizational practices evolve. Both
tacit experience accumulation and explicit knowl-
edge codication might precede the development
of organizational capabilities, at least in the context
of infrequent and heterogeneous processes, such as
corporate acquisitions (Zollo and Winter, 2002).
We view these two mechanisms as linked, in
that the effectiveness of knowledge codication as
a learning process depends, to some extent, on the
magnitude of accumulated experience (Levitt and
March, 1988). Yet they are also theoretically dis-
tinguishable because they assume different under-
lying behavioral and cognitive processes. Although
learning-by-doing occurs without explicit resource
commitment and intense cognitive effort, articu-
lating and codifying knowledge requires rms to
deliberately attempt to improve the odds of suc-
cess in future repetitions of the task. They can do
so only by dedicating time, money, and managerial
attention to grasp the causal mechanisms between
decisions, actions, and performance outcomes.
This argument suggests that the effects of the
process of knowledge codication, not necessarily
its outcomes, are of strategic relevance. Our argu-
ment focuses on the development and renement
of these tools, rather than on these tools’ usefulness
as repositories of collective memory (Cohen and
Bacdayan, 1994; Cohen et al., 1997) or as diffusers
of organizational knowledge (Nonaka, 1994; Non-
aka and Takeuchi, 1995). The process/outcome dis-
tinction is important for understanding the source
of sustainable advantages rms can derive from
their efforts to codify. Although the codication of
knowledge reduces the ability of rms to protect
their rents from imitation and replication (Winter,
1995) and might induce phenomena of supersti-
tious learning (Levitt and March, 1988), the supe-
rior understanding of the action–performance link-
ages derived from the creation of those tools will
not diffuse with the tools. Just as it is not suf-
cient to send a manual of explanations in order to
transfer superior practices (Szulanski, 1997), so it
will not be easy for competitors to reproduce the
performance of the initial codier even if they can
obtain access to the tools.
Our approach complements the ‘recombinatory’
(Kogut and Zander, 1992; Grant, 1996; Teece,
Pisano, and Shuen, 1997) and modular (Hender-
son and Clark, 1990; Clark and Fujimoto, 1991;
Sanchez and Mahoney, 1996) views of organiza-
tional capabilities, which emphasize the manipu-
lation of competence already residing within the
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1239
organization. Recombining, integrating, or ‘har-
nessing’ (Grant, 1996) current knowledge can
and should be distinguished from creating new
organizational competence in tasks that bear lit-
tle relationship with established rm activities.
The effectiveness of vicarious learning mech-
anisms for such tasks is limited by the very
nature of the organizational knowledge necessary
for their execution: sticky (Winter, 1995; Szulan-
ski, 1997), system-dependent (Winter, 1987), and
causally ambiguous (Lippman and Rumelt, 1982).
Finally, our approach offers the non-trivial advan-
tage of enhanced measurability, with respect to the
notions of combinative and architectural capabili-
ties because experience curves and the existence of
codication outputs can be easily quantied with
the appropriate methodology.
As do partnering, reengineering, and reorgani-
zation processes, acquisitions present a formidable
challenge for the rm attempting to develop a
specic capability in handling them. First, they
occur relatively infrequently and unpredictably,
thereby reducing a rm’s ability to accumulate
large amounts of ‘observations’ necessary to capi-
talize on learning-by-doing mechanisms (March,
Sproull, and Tamuz, 1991). Second, when they
do occur, they present themselves in highly het-
erogeneous forms and usually present a number
of unique challenges to be tackled (Haspeslagh
and Jemison, 1991). Third, this activity is inher-
ently causally ambiguous (Lippman and Rumelt,
1982), as the number, simultaneity, and interde-
pendence of the decisions and the actions entailed,
particularly in the context of the post-acquisition
integration phase, imply an endemic lack of clarity
with respect to their performance implications.
Under these conditions, the extent to which
acquiring rms codify the knowledge accumu-
lated through past experiences might be a nec-
essary precondition for the development of the
ability to manage the acquisition process. By cre-
ating and updating tools for executing the different
phases of the acquisition (i.e., negotiation, due
diligence, integration planning, and implementa-
tion), the acquiring rm might be able to form
and rene its understanding of the determinants
of performance outcomes. If it can, the quality of
its decisions and implementation steps should be
positively related to the degree to which it codi-
es knowledge from prior experiences in ad hoc
tools. Based on the above arguments, we propose
the following hypothesis:
Hypothesis 2: The higher the degree of knowl-
edge codication from previous acquisition ex-
periences, the better the economic performance
resulting from the focal acquisition.
Integration capability and task complexity
Supporting our hypothesis about the causal link
between knowledge codication and acquisition
performance is our observation that intense cog-
nitive efforts are necessary to develop and update
acquisition-specic tools that lower the degree of
causal ambiguity between decisions and perfor-
mance outcomes. If our argument is true, then
the effectiveness of the investments in deliber-
ate learning processes should increase, relative to
the process of accumulating tacit experience, at
increasing levels of task complexity (Zollo and
Winter, 2002). In other words, the more complex
is the task at hand, the higher its causal ambigu-
ity, and the more necessary it will be for the rm
to invest in deliberate learning efforts in order to
counteract the steeper barrier to the understanding
of cause–effect relationships.
A second argument in support of a possible
interaction of knowledge codication processes
with the degree of complexity of the organiza-
tional task has to do with the cognitive simplica-
tion entailed by the production of these tools. As
Gavetti and Levinthal (2000) show, investments
in forward-looking, ofine learning efforts should
reduce the cognitive complexity of the task at hand
by simplifying decision-making and facilitating the
coordination of the implementation subtasks.
This second point supplements the argument
Adler and Borys (1996) make in their analysis
of the conditions under which formalization might
be productive. The degree of formalization might
therefore represent a key to managing complex
contexts if it is enacted within a capability-building
approach, thereby enabling the achievement of
higher levels of understanding rather than coerc-
ing the actions of the individuals involved. The
former approach is more likely to be taken, how-
ever, when the learning challenge is greater. Easier
tasks might be more frequently approached with
coercive, bureaucratic attitudes.
In the acquisition context, the level of post-
acquisition integration of the acquired unit within
the acquiring rm is inherently tied to the com-
plexity of the organizational task. The higher the
level of integration is, the larger is the number
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1240 M. Zollo and H. Singh
of organizational units and functional departments
in both rms that need to coordinate and coop-
erate in order to achieve the desired structural,
operational, and cultural unity. The number, the
frequency, and the interdependence of decisions
and actions increase correspondingly, perhaps even
non-linearly, at increasing levels of intended inte-
gration between the two organizations.
Based on the above arguments, we present the
following hypothesis:
Hypothesis 3: The impact of the degree of knowl-
edge codication on performance will be strong-
er when the focal acquisition is managed with a
high, as opposed to a low, level of integration.
Performance implications of post-acquisition
integration strategies
Knowledge accumulation mechanisms do not,
however, solely determine the development of
acquisition-specic capabilities; this development
is obviously connected to, and is dependent
on, the type of integration approach selected
by the acquiring rm. Given the system-
dependent and causally ambiguous nature of the
organizational knowledge necessary to manage
an acquisition, acquirers will be able to develop
competence only in a fairly narrow knowledge
domain, which will likely correspond to the
management of a specic kind of acquisition
processes. When the characteristics of this process,
particularly of the post-acquisition integration
phase, change substantially, the acquirer should be
considered a novice, regardless of its accumulated
experience in substantially different types of
acquisitions. It thus has to start accumulating new
competencies specic to the new challenges it
faces. It is therefore very important to understand
the performance implications of post-acquisition
integration decisions (Haspeslagh and Jemison,
1991) and to relate them to the capability
development process described above.
As discussed in the section on ‘The perfor-
mance of corporate acquisitions,’ there appear to
be at least two key dimensions of the integra-
tion process: the level of organizational integration
between the two rms involved in the acquisi-
tion, and the extent to which the target’s resource
endowments, with particular emphasis on the tar-
get’s top management team, is replaced.
The level of integration
To the extent the acquired rm is integrated within
the structures and operations of the acquirer, a
number of outcomes are likely to occur. First, con-
sistent with established results on the effects of
organizational change on rm survival (Amburgey,
Kelly, and Barnett, 1993; Haveman, 1992, 1993),
a more extensive integration results in greater dis-
ruption of the pre-existing resources and routines
in both rms. This disruption is likely to lead
to declines in the performance of the combined
entity (Marks and Mirvis, 1985; Mirvis, 1985;
Schweiger, Ivancevich, and Power, 1987; Buono
and Bowditch, 1989, Astrachan, 1990; Empson,
A second negative consequence of the integra-
tion decision relates to its effect on the complex-
ity of the integration process. As argued above,
an extensive integration entails a large number
of highly interdependent and virtually simultane-
ous decision-making processes, involving increas-
ing levels of interaction between parts and func-
tions of the two organizations (Kitching, 1967;
Jemison and Sitkin, 1986; Pablo, 1994). Conse-
quently, extensive integration makes it harder for
acquirers to assess the performance outcomes of
the integration process and implies higher levels
of risk in the selection of the correct integration
approach, as well as in the implementation of
the selected approach (Pablo, Sitkin, and Jemison,
1996). Finally, high integration levels translate into
increasing explicit and hidden costs relative to the
expenses (e.g., training, lay-offs, information sys-
tems conversion), to the time and to the degree
of managerial attention (Ocasio, 1997) required to
design and implement the integration process.
Nonetheless, a higher level of integration be-
tween the two rms is necessary in order for the
acquirer to realize the potential value of the trans-
action (Datta and Grant, 1990; Haspeslagh and
Jemison, 1991; Shanley, 1994; Capron, 1999). In
particular, the positive performance implications of
the degree of resource relatedness (Rumelt, 1974,
1984; Chatterjee, 1986; Lubatkin, 1987; Singh and
Montgomery, 1987; Chatterjee et al., 1992) imply
that related acquisitions should be managed with
at least a minimum level of organizational integra-
Unfortunately, prior empirical work that has
attempted to link the level of integration to per-
formance has not yielded denitive results. Datta
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1241
and Grant (1990) did not nd statistically signi-
cant results for either their overall sample or their
subsample of related acquisitions, although their
sample of unrelated acquisitions did seem to ben-
et from lower levels of integration. In contrast,
Shanley (1994) found some evidence that positive
performance was related to the level of integration.
As our study focuses on horizontal and market
extension acquisitions, thereby excluding strate-
gic logics related to product diversication, we
expect the benets from economies of scale and
scope to emerge only when the operations of the
two organizations are integrated extensively, and
that these benets outweigh the negative impacts
of organizational disruptions, process complexity,
and implementation costs outlined above. Follow-
ing Datta and Grant (1990) and Shanley (1994),
we therefore propose the following hypothesis for
empirical testing:
Hypothesis 4: The higher the degree of integra-
tion of the acquired rm within the acquirer, the
better the economic performance of the acquisi-
The degree of resource replacement
We now consider the relationship between resource
replacement and acquisition performance. Of par-
ticular interest, given the attention it has received
in prior work, is the replacement of the top man-
agement team of the target rm. This variable
might be also considered a proxy, however, for a
more general construct of rm-wide replacement
of resources, such as brand names, distribution
channels, and physical assets, as an acquiring rm
opting for a quick, aggressive, integration process
is likely to replace other pre-existing resources
within the acquired business that it considers non-
We expect that when more extensive integration
is pursued, resource redundancy between the activ-
ities of the two organizations is more likely. Nev-
ertheless, given the decisions made on the degree
of integration, acquirers still have considerable lat-
itude in deciding how to achieve the desired level
of integration. For instance, the top management
team can be either retained and motivated to coop-
erate, or replaced with a new team sent from the
acquiring rm, regardless of how much autonomy
is given to the acquired organization. Also, if the
acquisition is primarily motivated by the access
to undervalued or underexploited assets, such as
brands or location, the decision on the retention of
top management is only loosely connected to the
one on the degree of integration of the productive
According to proponents of the ‘market for cor-
porate control’ hypothesis, the better team gains
control of the productive assets of the acquired rm
by displacing the less competent team (Manne,
1965; Jensen and Ruback, 1983). The acquiring
team, according to this argument, needs to believe
that it is more competent than the one currently in
place in the target rm. As reviewed in the section
on ‘The performance of corporate acquisitions,’
however, prior research suggests that replacing
the acquired rm’s top management will result in
reduced economic performance because it entails
the loss of human and social capital caused by
the departure of top executives. Empirical stud-
ies have found that managerial turnover reduces
acquisition performance (Cannella and Hambrick,
1993; Krishnan et al., 1997), lending support to a
negative sign of the causal link.
In an effort to nd an integrative solution to this
debate, we are interested in identifying a moder-
ating variable that might function as a ‘switch’ in
the sign of the relationship and that might have
been left out of the theoretical discourse so far.
To this end, an obvious moderator of the impact
of top management on performance is the quality
of the assets of the acquired rm. Replacing the
target’s top management might be connected with
enhanced performance if the acquired rm is char-
acterized by poor quality of its resources (and pre-
sumably of its pre-acquisition performance), since
the advantages of establishing a better management
team might outweigh the potential disruptions to
routines and motivation within the acquired orga-
nization. Conversely, replacing the management
team of better performing acquired rms is likely
to be detrimental to the performance of the com-
bined entity. If this is the case, then we can propose
the following hypothesis based on the interaction
between the replacement decision and the quality
of the pre-acquisition performance of the acquired
Hypothesis 5: The higher the pre-acquisition
performance of the acquired rm, the worse the
impact of the replacement of top management in
the acquired rm on the economic performance
of the acquisition.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1242 M. Zollo and H. Singh
Target Characteristics:
- Relative Size
- Market Relatedness
- Resource quality
- Top Management Replacement
- Level of Integration
Integration Capability
- Knowledge Codification
- Acquisition Experience
+ (H4)
+ (H5)
Post-acquisition Decisions:
Figure 1. Model and hypothesis
Figure 1 summarizes the causal relationships dis-
cussed and the research hypotheses advanced.
The research setting
We tested these hypotheses on a large sample of
acquisitions in the U.S. commercial banking indus-
try. This setting was deemed to be particularly well
suited for our research purposes for several rea-
sons. First, by holding the industry constant, we
ensured that the population of rms from which
we drew our sample was relatively uniform in the
environmental conditions that it faced. Of course,
the price paid was a reduction in the degree of
generalizability of our results, at least beyond the
banking sector. Second, this industry underwent
a period of rapid consolidation in response to
changes in regulation that allowed banks to cross
state lines in the early 1990s. This institutional
change created a sufciently large universe of
potential observations in a relatively compact time-
frame. Third, given the relevance of acquisition-
driven growth in the industry, it was easier to
nd rms that would grant us access and partici-
pate in our survey. Finally, banking has been the
most active industry in terms of acquisition vol-
umes since the beginning of the 1990s; its share
of the total domestic volume of acquisitions was
estimated to be about 25 percent during the rst
half of the decade.
The research design involved two phases: ini-
tial eldwork to gain a deeper understanding of
the integration process in this industry, and a
larger questionnaire-based sample study of post-
acquisition practices and performance. For the ini-
tial eld research, we obtained access to 12 banks,
all of which were active acquirers. We interviewed
45 decision-makers in these banks to obtain a bet-
ter understanding of how they handle integration
challenges and attempt to distill useful lessons
from their prior acquisition experiences. In ana-
lyzing the content of these interviews, we made
the following observations:
1. Most acquisitions made prior to 1990 were
managed as virtually autonomous afliates. Ac-
quired rms’ information systems were not
changed signicantly, and their top manage-
ment teams were typically not replaced.
2. During the 1990s, acquiring banks have increas-
ingly sought to obtain cost efciencies by inte-
grating the operations of the acquired bank
into their own, by standardizing products of
the combined organizations, and by converting
information systems.
3. Acquirers varied in how they managed acqui-
sitions not only longitudinally (increasing lev-
els of integration over time) but also cross-
sectionally. Some acquirers allowed the ac-
quired units to remain relatively autonomous,
and typically retained their top management.
Banc One was an example of this approach
to managing acquisitions. On the other hand,
equally experienced acquirers used substan-
tially different approaches to manage essentially
the same types of task by integrating and/or
replacing existing resources to a much greater
extent. Nationsbank (now Bank of America)
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1243
was considered one of the champions of this
more aggressive approach.
4. Finally, we were surprised to witness how
extensively some acquirers had codied the
integration process, and the large cross-sectional
variation that different acquirers exhibited along
this dimension. Experience levels seemed to
explain part, but not all, of this variation,
as we found several inexperienced acquirers
with highly sophisticated integration tools and
highly experienced ones with only average lev-
els of codication. Also, the development of
integration-specic tools went beyond the pre-
dictable, industry-specic challenges related to
the conversion of the IT systems, and spanned
issues of human resources management, cus-
tomer communication and retention, and gen-
eral project management systems.
The large sample study was then conducted in
1996, with a survey of the 250 largest bank
holding companies in the United States, cover-
ing more than 95 percent of the industry’s assets.
The asset size of the smallest institution was about
$400 million, which implies very rare acquisition
activity and very small transaction sizes (usually
one or two branches). Further extensions of the
sample to smaller institutions would probably have
resulted in very few responses because of the
scarcity of acquisitions in such rms’ histories, and
in low comparability among the observations in the
sample due to the large size differentials.
The survey consisted of two parts: The Acqui-
sition History Prole and The Acquiring Bank
Questionnaire. The Acquisition History Prole was
a list of all acquisitions conducted by the bank,
with basic information about each of them, such
as their asset size, the degree of market over-
lap, pre-acquisition protability, level of integra-
tion, and the replacement of the top manage-
ment team. The Acquiring Bank Questionnaire
described characteristics of the acquisition process
at the rm (rather than the transaction) level of
analysis, including the type and the time of cre-
ation of acquisition support tools such as integra-
tion manuals, system conversion manuals, product
mapping models, training packages, project man-
agement tools and other items. Of the 250 bank
holding companies contacted, 70 did not make
any acquisitions after 1985, and 16 were acquired
during the survey process. Of the remaining 164
banks, we obtained responses from 51, translating
into a response rate of 31.7 percent. This response
rate is satisfactory given the complexity of the sur-
vey and the involvement of top management in
responding to the survey.
The respondents to the survey were senior exec-
utives at the acquiring bank with direct experience
in the coordination of the acquisition process. Due
to variation in how rms were organized, we con-
tacted each potential respondent before the mailing
in order to identify the best respondent within
the bank. The respondents included the manager
responsible for corporate development and acqui-
sitions (26 institutions), the coordinator of post-
acquisition integration processes (14 institutions),
the CFO (eight institutions), and the CEO (in three
cases of smaller institutions).
Four responses had to be excluded from the
analysis because the data they supplied were seri-
ously incomplete. The remaining 47 institutions
had completed 577 acquisitions, for an average of
12.3 each. Standard mean comparison tests were
used to check for response bias. The responding
organizations were not signicantly different from
the original set of 250 organizations in terms of
return on assets, return on equity, or efciency
ratios, although they tended to be larger in terms
of asset size (p<0.05).
Dependent variable: performance
Acquisition performance is measured as the dif-
ference between return on assets (ROA) of the
acquiring bank 3 years after the acquisition vs.
the same measure 1 year before the acquisition.
The acquired banks in our study were very often
consolidated, from an accounting standpoint, into
the acquiring banks, leaving us unable to analyze
the target’s post-acquisition performance. Also, the
vast majority of acquired rms were privately held
community banks, whose accounting returns were
not publicly available even before the acquisi-
tion. We therefore resolved to utilize a measure
based solely on the acquiring bank’s pre- and post-
acquisition performance, and then collect a qual-
itative assessment of the target’s pre-acquisition
performance to use as a control variable.4
4See also our discussions in the section presenting the Results
for the robustness checks of our results, where we validate this
measure with a smaller subsample of observations for which we
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1244 M. Zollo and H. Singh
In addition to these measurement problems, our
interest in the performance implications of inte-
gration decisions and learning processes in the
acquiring rm meant it was appropriate to use
performance measures related to the acquirer. The
inuence of pre-acquisition performance and size
of the acquired unit are explicitly controlled for in
the multiple regression model. In order to control
for competitive conditions in the acquirer’s market,
we adjusted the acquiring bank’s return on assets
against the performance of its peers in the same
geographic area.5
The change in performance over time is then
expressed as
Change in ROA =(ROAi,t+3ROAc,t+3)
(ROAi,t1ROAc,t 1)
where ROAi,t+3and ROAi,t 1=return on assets of
acquiring bank iin years t+3 and t1 respec-
tively, and ROAc,t+3and ROAc,t1=average re-
turn on assets in the same geographic area as that
of the acquiring bank iat years t+3 and t1
The accounting data were collected from 1985
to 1997 with the use of three different databases
(Compustat, Compact Disclosures, and Moody’s)
in order to maximize the coverage of the bank-
ing sector. The coverage of the banking sector
(for both respondents and non-respondents) was
signicantly lower for the years prior to 1985.
Extending the data set would have implied a signif-
icant loss of comparability among the institutions
surveyed as well as consistency among the obser-
vations between the rst and the last years of the
period. Given the construction (3-year average) of
the dependent variable, the years 1985, 1995, 1996,
and 1997 were lost, thus restricting the period
of observation to acquisitions completed between
1986 and 1994.
could construct a weighted average of both banks’ ROA before
the acquisition. The two measures are essentially equivalent for
all practical purposes.
5Seven geographic areas in the United States (New England,
North Atlantic, South Atlantic, Mid-west, South, Rocky Moun-
tains, and Pacic) and one in Canada were used to benchmark
Explanatory variables
Knowledge codication is measured as the sum
of acquisition tools developed by the acquir-
ing rm at the time of the focal acquisition.
The tools are specic to different parts of the
acquisition process, including nancial evaluation,
due diligence, conversion of information systems,
human resources integration, and sales/product
integration. The information was gathered through
the Acquiring Bank Questionnaire, which asked
whether the following items were developed and,
if so, when they were developed:
Documents/Manuals: Due diligence checklist,
Due diligence manual, Systems conversion man-
ual, Afliation/integration manual,6Systems
training manual,7Products training manual.8
Quantitative Models: Financial evaluation, Staff-
ing models, Product mapping,9Training/Self-
training packages, Project management.10
Acquisition experience is computed as the num-
ber of acquisitions completed by the acquiring
rm before the focal acquisitions. The Acquisition
History Prole collected the list of all the acquisi-
tions completed by the responding institution since
founding or since a merger of equals. The oldest
acquisitions in the data set were completed in 1968
(by Banc One and Crestar Bank). Although the
analysis is based on the observations between 1986
and 1994 (see above), the History Prole allowed
us to construct the complete stock of prior acquisi-
tion experience for each of the observations in the
analysis. The measurements of the two capability-
building mechanisms, as well as of the dependent
variables, are therefore comparable across rms.
Integration was measured with a single scale
collected with the Acquisition History Prole in-
strument. For each acquisition they completed
since their founding, respondents answered the fol-
lowing question:
6Such manuals describe all the procedures necessary to
accomplish the desired level of integration between the two
organizations. They usually cover issues such as human
resources, accounting, audit, and CRA.
7These manuals describe how to train the DP users at the
acquired company. They are ‘train-the-trainer’ tools.
8These manuals describe how to train the sales force at the
acquired company.
9These packages allow thorough comparison of the features of
the acquired bank’s products with those of the acquirer.
10 These models assign tasks, requirements, and deadlines, allow-
ing careful planning and control of complex projects.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1245
Integration. To what extent were the systems,
procedures and products aligned or centralized?
Possible answers were: ‘0’ (few or no features
were aligned or centralized), ‘1’ (if only selected
systems, procedures or products were aligned or
centralized), ‘2’ (many but not all systems, proce-
dures, and products were aligned or centralized),
and ‘3’ (all systems, procedures, and products were
completely integrated).
Replacement was measured with a similar four-
point scale, where respondents answered the fol-
lowing question:11
Change. To what extent has the executive lead-
ership of the acquired bank been changed after
the acquisition?
Alternative answers were: ‘0’ (no substantial
change), ‘1’ (some changes), ‘2’ (many changes),
and ‘3’ (virtually all the top management team was
The research design called for limiting the varia-
tion in the degree of relatedness between the two
organizations to the geographic dimension. The
construct was therefore measured with a dummy
variable identifying either horizontal acquisitions
(or ‘in-market’ in the banking terminology), coded
as ‘1’, or market extension acquisition (or ‘out-
market’), coded as ‘0’. This measure is a good
proxy for market relatedness in the banking indus-
try, given the importance of geographic location
as a key competitive factor and the importance
of rationalizing the branch network in order to
create value from acquisitions through cost ef-
ciencies. In terms of value creation mechanisms,
in-market acquisitions generally prioritize cost ef-
ciencies driven by economies of scale, whereas
‘out-market’ acquisitions tend to rely more on
cross-selling opportunities and economies of scope
since many of the cost-efciencies derivable from
11 The use of the word ‘replacement’ was deemed too negative.
We therefore substituted the word ‘change’ without signicantly
altering the meaning of the question. Note that the use of the
phrase ‘has been changed,’ rather than ‘has changed,’ implies an
active role of the acquirer in substituting the top management of
the acquired bank. We are likely to capture, therefore, deliberate
lay-offs by the acquirer as opposed to retention problems.
the rationalization of the two branch networks are
not available.
Resource quality
In order to isolate the effect of the resource
replacement variable, we assessed the pre-acquisi-
tion quality of the acquired rm’s resource endow-
ment. The construct was measured by asking about
the performance level of the target bank prior to
the acquisition. The scale anchors were: ‘2’ (the
acquired institution was bankrupt), ‘1’ (it was a
poor performer), ‘0’ (it was an average performer),
+1’ (it was a good performer) and ‘+2’ (it was
an outstanding performer).
Other controls included the asset size of the
acquiring rm, the relative asset size of the acquir-
ed rm with respect to the acquirer, and the num-
ber of acquisitions completed by the responding
bank during the same year of the focal acquisition.
Construct validity
We tested the validity of our measures of integra-
tion, replacement, relatedness, and resource qual-
ity using multiple item scales we developed for
a subsample of 57 acquisitions. For this subsam-
ple, we had multiple indicators for resource relat-
edness (8 items), resource quality (11), manage-
rial replacement (9), and the degree of integration
both as product and process alignment (8) and
as functional centralization (7, with a total of 15
items for the integration construct). Comparisons
of the means between this subsample and the entire
database of acquisitions did not indicate any bias
for these constructs. We used three tests to check
the validity of our measures; (1) Cronbach alphas
of the multiple items, (2) correlation between the
scale used in the study and the sum of the z-scores
of multiple items, and (3) correlation between the
scale used in the study with the main factor
extracted from the multiple items.
Results indicate that the measures used in the
empirical analysis are generally valid representa-
tions of the underlying constructs utilized in the
theoretical treatment of acquisition performance.
For our measure of top management replacement,
the Cronbach alpha, correlation of the scale with
the sum of the z-scores for the multiple items,
and the correlation between the scale and the
main factor are 0.826, 0.606 (p<0.01), and 0.549
(p<0.01), respectively, all of which indicate that
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1246 M. Zollo and H. Singh
our scale for the replacement of top management
is valid. For the degree of integration, the same
statistics are 0.950, 0.521 (p<0.01), and 0.542
(p<0.01), all of which indicate that our integra-
tion scale is valid. We also validated our resource
quality construct (Cronbach alpha =0.853; corre-
lation with the sum of the z-scores for the multi-
ple items =0.463, p<0.01; correlation with the
main factor =0.482, p<0.01). The only item
that was not validated was the approximation of
resource relatedness with the single measure of
market overlap (in-market vs. out-market acqui-
sitions). Nonetheless, the single measure does cor-
relate strongly with a more focused set of items
related to the degree of overlap in terms of branch
network and customer base (0.520, p<0.01).
Consequently, we replaced the general notion of
resource replacement with the narrower concept
of market relatedness in our interpretations of the
results of this analysis.
The model
The model being tested in this study is specied
as follows:
Change in ROA =a+bintegration
creplacement +dcodication
+fcodication ×integration
+greplacement ×quality
+controls +ε
The error term is distributed according to the
standard normality assumptions.
The estimation method used is ordinary least
squares. All variables utilized to construct the
interaction terms were standardized so as to elim-
inate the initial multicollinearity problem in the
estimated model. With this correction, the max-
imum variance ination factor (VIF) across the
covariates was 2.591, which is signicantly below
the rule of thumb of 10 used to detect multi-
collinearity problems (Neter, Wasserman, and Kut-
ner, 1985). We also checked the stability of the
coefcients to different specications of the model,
dropping one of the post-acquisition decisions or
one of the (highly correlated) resource character-
istics of the acquired rm, and found no impor-
tant variations to the magnitude and statistical
signicance of the coefcients. Four observations
were identied as outliers (>3 S.D.) and were
excluded from the analysis. The only other vio-
lation of standard normality assumptions that we
could nd was related to a possible correlation of
the error terms stemming from the multiple obser-
vations for each responding institution. We dealt
with this concern through a specic set of analyses
reported in the next section.
Table 1 reports descriptive statistics and the cor-
relation matrix for the data used in this study.
Consistent with the prior literature on mergers and
acquisitions, we nd that the mean for the per-
formance variable (0.004) is not statistically dis-
tinguishable from zero. The correlation table indi-
cates that the dependent variable is signicantly
correlated with virtually all of our explanatory
variables, with the notable exception of acquisition
experience. Many of our explanatory variables are
also correlated with each other. We thus used a
multivariate analysis to identify the net inuence
of each variable on acquisition performance.
The results of the regression analysis of the
model described earlier are reported in Table 2.
The six nested models presented allow the effect
of each group of variables on acquisition perfor-
mance to be identied. The models t with the data
reasonably well, as shown by the strongly signi-
cant F-statistics (p<0.001) and by the increasing
adjusted R2statistic (0.165 in the full model). The
incremental F-statistic (not reported) is statisti-
cally signicant in each model, with the excep-
tion of the one introducing the level of acquisition
experience. The two focal post-acquisition deci-
sions appear to impact the variation in acquisition
performance more strongly than any of the other
sets of explanations. We enter the two decisions
simultaneously in Model 3 because of their rel-
atively high correlation ratio; a sequential entry
would not have changed the results in a substan-
tive way, but would have shown a biased esti-
mate of the coefcient of the rst decision entered,
since part of the variation of the second decision
would have been picked up. The organizational
learning variables, as well as the pre-acquisition
resource characteristics, which we entered sepa-
rately because of their theoretical relevance, show
mixed results, with only one of the two variables
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1247
Table 1. Correlation matrix
Variables Avg. Std 1 2 3 4 5 6 7 8 9 10
1 Acquisition perf. ch. in ROA (3 years) 0.004 0.3518
2 Acquirer’s size 23.137 23.08 0.084
3 Relative acquisition size 6.108 11.45 0.013 0.075
4 Simultaneous acquisitions 3.589 2.836 0.235 0.481 0.223
5 Resource quality 0.017 1.06 0.070 0.048 0.052 0.051
6 Market relatedness 0.61 0.49 0.040 0.178 0.080 0.144 0.201
7 Level of integration 2.635 0.703 0.115 0.089 0.090 0.172 0.213 0.396
8 Degree of replacement 1.76 1.28 0.230 0.067 0.019 0.210 0.309 0.352 0.417
9 Acquisition experience 11.34 10.17 0.026 0.502 0.090 0.515 0.036 0.170 0.118 0.057
10 Knowledge codication 4.877 3.676 0.146 0.436 0.052 0.360 0.175 0.026 0.078 0.109 0.455
11 Codication ×Integration 0.173 1.155 0.143 0.023 0.072 0.047 0.063 0.009 0.154 0.111 0.066 0.176
Pearson’s correlation. Bold numbers are signicant at the 5% level; bold and italic ones are signicant at the 1% level.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1248 M. Zollo and H. Singh
Table 2. Acquisition performance: transaction level of analysis
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Acquirer’s size 0.031 0.044 0.064 0.054 0.127 0.118
Relative acquisition size 0.072 0.065 0.080 0.077 0.096 0.072
Simultaneous acquisitions 0.298∗∗∗ 0.308∗∗∗ 0.129 0.149 0.155 0.113
Target quality 0.136∗∗ 0.18∗∗∗ 0.179∗∗∗ 0.221∗∗∗ 0.227∗∗∗
Market relatedness 0.001 0.062 0.064 0.068 0.061
Post-acquisition decisions
Level of integration 0.181∗∗ 0.174∗∗ 0.176∗∗ 0.185∗∗
Degree of replacement 0.344∗∗∗ 0.339∗∗∗ 0.338∗∗∗ 0.336∗∗∗
Integration capability
Acquisition experience 0.037 0.098 0.081
Knowledge codication 0.198∗∗ 0.207∗∗∗
Codication ×Integration 0.142∗∗
F-statistic 6.024∗∗∗ 4.544∗∗∗ 5.976∗∗∗ 5.234∗∗∗ 5.455∗∗∗ 5.506∗∗∗
Adjusted R20.062 0.072 0.133 0.129 0.150 0.165
N228 228 228 228 228 228
Ordinary least squares estimation. Standardized beta coefcients: signicant at the 1% (∗∗∗), 5% (∗∗) or 10% () level.
Dependent variable: change in ROA 3 years after the acquisition vs. the year before, minus same variation in local competitors’
signicantly inuencing performance. The stabil-
ity of the coefcients across the models, as well
as customary controls of VIF and tolerance ratios,
shows that multicollinearity problems, which could
arise from high correlations among groups of vari-
ables, are not present in this sample.12
The analysis indicates that Hypothesis 2, which
posits a positive relationship between knowledge
codication and performance, is strongly sup-
ported (p<0.01). In addition, the interaction
between knowledge codication and the level of
integration is statistically signicant at the 5 per-
cent level, supporting Hypothesis 3. The fact that
the creation of these tools shows increasing inu-
ence on acquisition performance at increasing lev-
els of integration (i.e., task complexity) empirically
supports our argument about the role of deliberate
learning efforts in the development of integration
In contrast, the accumulation of tacit knowl-
edge through acquisition experience turns out to
be a non-signicant predictor of performance, fail-
ing to support Hypothesis 1. This nding conrms
the mixed results of the received literature on the
performance implications of accumulating acqui-
sition experience. The data analyzed suggest that
12 Particularly important is the stability of the coefcients across
Models 4, 5, and 6 despite the signicant correlation between
acquisition experience and knowledge codication.
in the context of relatively infrequent, heteroge-
neous, and causally ambiguous tasks,13 organiza-
tions develop competence primarily by articulating
and codifying knowledge derived from previous
acquisition experiences. Simple exposure to acqui-
sition processes does not seem to sufce.
Hypothesis 4, suggesting that the level of inte-
gration is positively associated with changes in
performance, is supported at the 1 percent sig-
nicance level. Regarding the performance impli-
cations of top management replacement, we rst
note that the direct effect is negative and signi-
cant at the 1 percent level, lending further support
to the ‘organizational disruption’ view (Cannella
and Hambrick, 1993) vs. the ‘market for corpo-
rate control’ perspective (Manne, 1965; Jensen and
Ruback, 1983). The strength of the result was
surprising, given theoretical arguments supporting
both a positive and a negative impact for this deci-
sion. However, this result could be explained by
the presence of a larger number of good perform-
ing targets in the sample analyzed. According to
Hypothesis 5, in fact, the pre-acquisition perfor-
mance levels of the acquired rm should moderate
the relationship between replacement and acquisi-
tion performance.
13 Note that the assessment of acquisitions as infrequent and
homogeneous tasks needs to be viewed in comparative terms
with respect to more standard organizational activities, such as
operating or administrative tasks.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1249
In order to probe this, we added the interaction
term between replacement and the resource qual-
ity assessment of the acquired bank. In addition,
we split the dataset into three subsamples by sep-
arating our observations into low (i.e., response =
2or1), average (response =0), and high
(response =+1or+2) resource quality. Table 3
reports the results of OLS estimates with the full
model including the interaction term, as well as for
the three subsamples. As Hypothesis 5 proposes,
the interaction term is strongly signicant and neg-
atively related to performance, indicating that top
management replacement is increasingly correlated
with poorer performance when the resource quality
of the acquired unit increases. More interestingly,
though, the strongest negative effect of the replace-
ment decision is found for observations with an
average quality of resources, though it would be
logical to expect a stronger negative effect for
observations with the highest quality of resources.
At low quality levels, the replacement decision
takes on a positive sign, as per the ‘market for
corporate control’ hypothesis, but does not reach
statistical signicance.14
14 These results are conrmed by a Chow test analysis on
the replacement coefcient as well as on the entire model.
For the entire model, the average level of resource quality
changes the slope signicantly vis-`a-vis the other two levels
One possible explanation for this non-linearity
in the TMT replacement coefcients could be pro-
vided by the fact that the uncertainty about the per-
formance implications of replacement is highest at
intermediate levels of pre-acquisition performance.
Also, the model seems to t the data very well for
low and average performing targets, but not for
high-quality ones. This might be a consequence of
the paucity of degrees of freedom (only 55 obser-
vations) and a possible overspecication of the
model. It might also be, though, that some of the
prescriptions valid for the acquisition of ‘normal’
targets might not transfer to high-quality ones. Cost
efciencies from the integration of high-quality
targets, for example, might be effectively counter-
balanced by higher hazards of disruption of supe-
rior routines. In terms of learning, the advantages
(F=6.595, p<0.001), the high-quality model adds some
additional shift but is only marginally signicant (F=1.661,
p<0.10), whereas the low-quality coefcients do not differ
from the others in a statistically signicant way (F=1.235).
Similarly, the single coefcient of TMT replacement shows the
strongest negative impact in the interaction with the average
quality dummy (t=−5.134, p<0.0001), followed by the high-
quality one (t=−2.113, p<0.05), with the one with the low-
quality resources indeed switching sign but failing to reach
statistical signicance (t=+0.466).We wish to acknowledge
that the results on the interaction effect between replacement
and resource quality were the consequence of one reviewer’s
insightful comments and generous advice, for which we are
particularly grateful.
Table 3. Test for interaction between TMT replacement and target quality
Interaction term Low quality Avg. quality High quality
Acquirer’s size 0.061 0.031 0.152 0.031
Relative acquisition size 0.061 0.032 0.034 0.193
Simultaneous acquisitions 0.044 0.022 0.041 0.064
Target quality 0.151∗∗
Market relatedness 0.082 0.122 0.133 0.183
Post-acquisition decisions
Level of integration 0.309∗∗∗ 0.397∗∗∗ 0.365∗∗∗ 0.091
Degree of replacement 0.401∗∗∗ 0.144 0.621∗∗∗ 0.359∗∗
Integration capability
Acquisition experience 0.095 0.140 0.037 0.013
Knowledge codication 0.151∗∗ 0.2260.014 0.153
Codication ×Integration 0.168∗∗∗ 0.255∗∗ 0.081 0.263
Replacement ×Quality 0.228∗∗∗
F statistic 7.131∗∗∗ 2.61∗∗∗ 5.963∗∗∗ 1.253
Adjusted R20.221 0.148 0.314 0.040
N 238 83 98 55
Ordinary least squares estimation. Standardized beta coefcients: signicant at the 1% (∗∗∗), 5% (∗∗) or 10% () level.
Dependent variable: change in ROA 3 years after the acquisition vs. the year before, minus same variation in local competitors’
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1250 M. Zollo and H. Singh
of knowledge codication might be reduced out-
side the more routinizable case of restructuring
poorly performing targets.
Compared to post-acquisition decisions, and
knowledge codication, the characteristics of the
pre-acquisition resources of the target, resource
quality, and market relatedness, show weaker ex-
planatory power, with the market relatedness mea-
sure showing no statistically signicant effect. This
nding is puzzling from a theoretical standpoint,
because the potential for economies of scale should
be signicantly superior for horizontal acquisi-
tions than for market extension ones, as the over-
lap of two branch networks typically allows ef-
ciency gains from rationalization. One interpreta-
tion might be that acquirers can create or destroy
value equally well through cost rationalizations,
typically prioritized in ‘in-market’ acquisitions, or
through revenue enhancement processes, which
become the priority in market extension acquisi-
Resource quality consistently impacts perfor-
mance negatively, indicating that the acquisition
of well-performing targets is less likely to enhance
acquirers’ performance than is the acquisition of
poorly performing ones. This result can be inter-
preted in two ways: rst, in terms of directionality
of the knowledge ows between the acquiring and
the acquired rm. Consistent with Capron’s (1999)
results on the performance implications of resource
redeployment to and from the target, our nding
can be interpreted as showing that the transfer of
resources and capabilities from the acquirer to the
target (i.e., the cases in which the target quality is
low) outperforms the opposite mechanism, through
which the acquirer ‘learns’ from the (highly per-
forming) acquired entity. The second interpretation
refers to increasing levels of resistance to change
that can be expected to occur as the levels of pre-
acquisition performance of the acquired rm grow.
The better the acquired rm performed before the
acquisition, the stronger the condence its man-
agers will have in the superior quality of its pro-
cesses and the less willing they will be to accept
the changes required to align processes and proce-
dures across the two organizations. Interestingly,
this result also replicates ndings in prior nance
literature on U.S. bank mergers studying short-
term stock price reactions to merger announce-
ments (Hawawini and Swary, 1990).
None of the other variables entered as con-
trols in the model—acquirer’s size, acqui-
sition relative size, and the frequency of
simultaneous acquisitions—signicantly inu-
ences performance, further suggesting that the
variables considered in our theoretical discussion
are meaningful and relevant to the explanation of
acquisition performance.
Robustness of results
As we noted in the previous section, the error terms
in our estimates of the model may not be inde-
pendently distributed because we have multiple
observations from the same responding institution.
In order to address this problem, we replicated
the analysis by aggregating the data in two dif-
ferent ways: the rm/year level of analysis, where
all acquisitions in the same year by the same
rms were averaged, and the rm level of anal-
ysis, where all the acquisitions completed by the
same rm were aggregated. Weighted least squares
estimations (where weights are assigned by the
number of acquisitions completed by the same
bank over the period of observation) at these two
levels of analysis yield results that are consistent
with the ones described above, in spite of the sig-
nicantly lower number of degrees of freedom.15
The result offers further evidence that the esti-
mated model accurately characterizes this sample
(see Table 4). Using rm-level dummies to control
for rm effects was not appropriate, as they would
have picked up the measurement errors of our main
theoretical variables (experience accumulation and
knowledge codication). Also, although rm con-
trols might alleviate this problem, the aggregation
of the data eliminates it by making only one obser-
vation available for each respondent.
Another issue of concern was related to our
measurement of the dependent variable, which did
not include the ROA of the acquired bank before
the acquisition, while the post-acquisition value
included the acquired unit. In response to one
of the reviewers’ suggestions, we collected pre-
acquisition ROA data for all the acquired banks
we could nd (only 79 complete observations) and
15 Our sample size dropped from 47 to 31 banks because a lot of
institutions completed their acquisitions only in the last 3 years
of the observed period (1985–97), and therefore are missing
from the analysis. Also, the aggregation process was restrictive,
in the sense that any bank with one missing value in any one of
the acquisitions completed was removed from the analysis.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1251
Table 4. Acquisition performance: rm level of analysis
Variable Model 1 Model 2 Model 3 Model 4
Resource-based factors
Quality of assets (avg.) 0.259 0.310 0.315 0.549∗∗∗
Market relatedness 0.250 0.013 0.031 0.109
Post-acq. decisions
Integration 0.559∗∗∗ 0.546∗∗∗ 0.518∗∗
Replacement 0.292 0.259 0.292
Integration capability
Acquisition experience 0.066 0.068
Knowledge codication 0.515∗∗∗
F-statistic 3.0684.368∗∗∗ 3.408∗∗ 5.364∗∗∗
Fimprovement 3.0684.853∗∗ 0.130 9.546∗∗∗
Adjusted R20.118 0.303 0.280 0.458
N31 31 31 31
Weighted least squares estimations. Standardized beta coefcients. Signicant at the 1% (∗∗∗), 5% (∗∗ ) or 10% () level.
Dependent variables: avg. change in ROA over the period of analysis.
Note: Due to the low N, variables that had no signicant effect in previous analyses were omitted in order to save d.f.
constructed a measure of acquisition performance,
which included the asset-weighted average of ROA
for the two banks 1 year before the acquisition.
The correlation of the resulting measure with the
one utilized in the reported analysis is 0.972 (p<
0.0001), conrming that the two approaches pro-
vide almost identical estimates of the dependent
This paper has discussed how post-acquisition
decisions and capability-building processes affect
the economic performance of corporate acquisi-
tions. We proposed a knowledge-based perspective
of acquisitions, which builds on the intuition that in
order to enhance acquisition performance acquir-
ing rms need to not only select the appropriate
mix of integration decisions but they have to simul-
taneously develop the organizational capability to
implement it. We drew upon multiple theoreti-
cal traditions, using resource-based, process-based,
and evolutionary economics arguments to enrich
the existing literature on the knowledge-based
view of the rm (Kogut and Zander, 1992; Grant,
1996). One crucial insight, which seems to be sup-
ported by our data, is that rms develop collective
competence by not only accumulating experience
but also investing time and effort in activities that
require greater cognitive effort in order to pro-
duce enhanced awareness of action–performance
linkages.16 Firms learn directly by articulating and
codifying the lessons they learned from previous
experiences, even if they might not be aware of the
positive learning spillovers from these activities.
At an extreme, the benet in creating and ne-
tuning acquisition-specic tools might lie more in
the learning achieved through the creative process
itself than in the use of the outputs as coordination
and implementation support devices.
The results of the analysis also suggest that the
‘process view’ of acquisitions (Jemison and Sitkin,
1986; Haspeslagh and Jemison, 1991; Pablo, 1994;
Pablo et al., 1996), which emphasizes the role
of the integration phase is relevant to consider
in understanding the performance of the entire
acquisition process. Although the results of prior
attempts to relate the level of integration to perfor-
mance are equivocal, our nding suggests that, at
least in the banking industry, which has a trend of
efciency-driven consolidation, the benets from
cost efciencies gained through higher levels of
integration might be greater than the costs inher-
ent to the integration process (e.g., routine and
competence disruptions, increased process com-
plexity, and hidden implementation costs). Thus,
in this setting, the negative consequences typically
16 As Zollo and Winter (2002) noted, deliberate learning efforts
are not to be viewed as systematically superior mechanisms for
capability development. In tasks more frequent, less heteroge-
neous and less causally ambiguous than the type studied, the
costs connected with knowledge articulation and codication
processes may overcome the benets rms derive from them.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1252 M. Zollo and H. Singh
attributed to post-acquisition integration processes
within the human resources management and orga-
nizational behavior literature do not systematically
occur (Marks and Mirvis, 1985; Mirvis, 1985;
Schweiger et al., 1987; Buono and Bowditch 1989;
Astrachan, 1990).
In addition, we tested for the direct effect of the
replacement of top management in the acquired
rm on acquisition performance, as well as the
inuence that the pre-acquisition performance of
the acquired rm has on the strength of the rela-
tionship. Our results show that the main effect
of the replacement decision in the context stud-
ied is negative and signicant, and that pre-
acquisition performance impacts the relationship
in the hypothesized sense but in a non-linear way.
The negative impact of top management replace-
ment on performance is maximum at intermediate
levels of pre-acquisition performance, rather than
at high levels. Moreover, the sign of the impact
switches to positive with low-performing targets,
but does not reach statistical signicance. Taken
together, these results conrm the value of search-
ing for an integrative solution to the debate in the
literature on the view of acquisitions as policing
mechanisms for agency problems in the ‘market
for corporate control.’ Results seem to indicate that
this view might be applicable only to the case of
under-performing targets, but applying its tenets to
cases of average (as well as superior) performers
might lead to lower performance for the combined
The knowledge-based variables in the model
show interesting effects. The degree of codication
has a strong and positive inuence on acquisition
performance; as the rst of its type, this nding
merits subsequent research to test for its gener-
alizability. In contrast, the impact of experience
accumulation is non-signicant. This latter result
adds to a series of mixed ndings on experience
accumulation in these types of tasks. Learning
curve effects in the context of highly infrequent
and heterogeneous events such as those studied17
might be heavily attenuated as the hazards of erro-
neous generalization from the lessons it learned in
past contexts to seemingly similar but inherently
different ones are correspondingly high (Cormier
17 Again, note that the characterization of acquisitions as infre-
quent and heterogeneous tasks is made in comparative terms,
vis-`a-vis normal operating or administrative activities, the typi-
cal subject of received studies in learning curves.
and Hagman, 1987; Cohen and Bacdayan, 1994;
Haleblian and Finkelstein, 1999; Levitt and March,
Importantly, and central to our arguments on the
co-evolution of integration decisions and capabil-
ity building processes, the interaction between the
degree of codication and the level of integration
positively and signicantly inuences acquisition
performance. At increasing levels of complexity,
the benets of explicitly extracting lessons learned
from previous experiences appear to exceed the
costs connected to codication activities (e.g.,
investment in time, effort, and managerial atten-
tion). This result is important because it is direct
evidence of the relationship between acquisition
capabilities and the management of more com-
plex integration decisions, and may explain why
a large number of integrations (even simpler ones)
are not successful. Recent related research is exam-
ining the role of post-acquisition organizational
decisions in inuencing product performance in
knowledge-intensive acquisitions, another domain
replete with instances of post-acquisition failure
(see, for example, Puranam, Singh, and Zollo,
We also introduced the type of acquisition (hor-
izontal or market extension) as an important con-
trol variable in the analysis. The lack of signi-
cant performance implications for this variable is
interesting. Ex ante, one could argue that acquir-
ing a competitor in the same geographic area
would create higher potential for efciency-driven
cost reductions. Such acquisitions might, however,
require more complex integration efforts in terms
of the number of potential overlaps of resources
and activities across the organizations and the con-
sequently large array of simultaneous, interdepen-
dent decisions and actions necessary to accom-
plish this integration. Therefore the characteristics
of pre-acquisition resources might not necessarily
predict post-acquisition performance. Instead, the
set of post-acquisition decisions about manipulat-
ing those resources, the capability to do so that
the acquiring rm eventually develops, and the t
between these two factors, seem to matter most.
It is important to note, however, that our sam-
ple deliberately excludes product extensions and
unrelated acquisitions, so the range of variation on
acquisition type is less extensive than it is in most
other studies.
This study has other limitations. It is a single
industry study, focusing on U.S. bank mergers.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1253
Its applicability to other industries and other geo-
graphic and institutional contexts needs to be,
therefore, closely examined. This problem might
be particularly relevant for the generalizability of
the performance implications of integration deci-
sions. The results related to the capability-building
mechanisms might be more safely extended to
other types of acquisitions, and in fact to similarly
complex organizational tasks, such as alliances and
internal restructuring processes, as long as they
maintain comparable levels of (in)frequency, het-
erogeneity and causal ambiguity. Importantly, the
study is also based on a limited denition of acqui-
sition performance, which emphasizes the varia-
tion in performance of the acquiring rm. The
measurement of the dependent variable, based on
accounting data, could also be effectively corrob-
orated with other proxies.
This study of acquisitions attempts to bridge
and integrate different theoretical approaches to the
highly visible phenomenon of corporate acquisi-
tions. In spite of the economic relevance of the
phenomenon, when rms turn to academia for
some guidance on how to improve their chances
of creating value from their investments, they are
typically met with a set of highly segmented rec-
ommendations. Finance scholars point to the fact
that acquisitions on average do not create abnor-
mal returns for the acquirers, raising questions of
acquirers’ motives for engaging in these trans-
actions. Strategy scholars are slightly more opti-
mistic, distinguishing between more sensible (i.e.,
related) and less sensible (i.e., unrelated) types of
investments. Finally, scholars in organization stud-
ies emphasize the hardships connected with the
effective management of the integration phase, the
disruption of existing resources and competencies,
and the loss of managerial and operational talent.
We hope this study will help to signal the
advantages of leveraging different theoretical per-
spectives in offering managers a more clearly
dened and useful account of the conditions under
which competitive advantage can be gained or
destroyed in acquisition activities. Acquisitions,
like any other challenging organizational task, can
be effectively managed in a consistently value-
creating way, if the conditions enabling perfor-
mance enhancement are correctly identied and
exploited. Even more importantly, rms seem to
be capable of developing specic capabilities that
allow them to improve their chances of success
over time. The data analyzed in this study show
that deliberate learning processes, as opposed to
semi-automatic (e.g., learning-by-doing) ones, play
an important role in predicting acquisition perfor-
mance, providing some indications of the way an
acquisition capability may develop.
More studies will be necessary in order to test
our hypotheses in different contexts and to achieve
a more ne-grained appreciation of the condi-
tions under which distinct integration strategies
work and how integration capabilities develop.
We believe that the results of the analyses pre-
sented above can guide future scholars in promis-
ing directions toward increasing understanding of
the antecedents of merger performance. This work
also has implications, more broadly, for the cre-
ation and evolution of organizational capabilities.
Further research on the processes used by rms
to develop capabilities in other contexts, such as
restructuring process, strategic alliances and new
product development, will strengthen our under-
standing of these important phenomena. In addi-
tion, such research will provide additional settings
in which to apply, validate and extend the knowl-
edge and capability based view of the rm.
Generous funding from the Sloan Foundation and
support from the Wharton Financial Institutions
Center, the Mack Center for Technological Inno-
vation, and the R&D Department at INSEAD, are
gratefully acknowledged. Many useful comments
on previous versions of this paper, titled ‘Post-
acquisition Strategies, Integration Capability, and
the Economic Performance of Corporate Acquisi-
tions,’ were received from Sea-Jin Chang, Philippe
Haspeslagh, David Jemison, Prashant Kale, John
Kimberly, Dan Levinthal, Phanish Puranam, Jeff
Reuer, Jose Santos, Anju Seth, Gabriel Szulan-
ski, and Sid Winter. All remaining errors are our
Adler P, Borys B. 1996. Two types of bureaucracy:
enabling and coercive. Administrative Science Quar-
terly 41: 61–89.
Agrawal A, Jaffe JF. 2000. The post merger performance
puzzle. In Advances in Mergers and Acquisitions,
Gregory A, Cooper C (eds). JAI: New York; 7–41.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1254 M. Zollo and H. Singh
Agrawal A, Jaffe JF, Mandelker GN. 1992. The post-
merger performance of acquiring rms: a re-
examination of an anomaly. Journal of Finance 47(4):
Amburgey TA, Kelly D, Barnett WP. 1993. Resetting the
clock: the dynamics of organizational change and
failure. Administrative Science Quarterly 38: 51 –73.
Anand J, Singh H. 1997. Asset redeployment, acquisition
and corporate strategies in declining industries.
Strategic Management Journal, Summer Special Issue
18: 99–118.
Astrachan JH. 1990. Mergers, Acquisitions, and Employee
Anxiety: A Study of Separation Anxiety in a Corporate
Context. Praeger: New York.
Baker G, Montgomery C. 1974. Conglomerates and LBO
associations: a comparison of organizational forms.
Working Paper, Harvard Business School.
Barney JB. 1986. Strategic factor markets: expectations,
luck, and business strategy. Management Science
32(10): 1231–1241.
Barney JB. 1988. Returns to bidding rms in mergers and
acquisitions: reconsidering the relatedness hypothesis.
Strategic Management Journal, Special Issue 9:
Baum JAC, Ginsberg A. 1997. Acquisition experience
and protability: exploring the value of learning by
doing. Working paper, New York University.
Buono AF, Bowditch JL. 1989. The Human Side
of Mergers and Acquisitions. Jossey-Bass: San
Francisco, CA.
Bruton GD, Oviatt BM, White MA. 1994. Performance
of acquisitions of distressed rms. Academy of
Management Journal 37: 972 –989.
Cannella AA Jr, Hambrick DC. 1993. Effects of
executive departures on the performance of acquired
rms. Strategic Management Journal, Summer
Special Issue 14: 137–152.
Capron L. 1999. The long-term performance of horizontal
acquisitions. Strategic Management Journal 20(11):
Chatterjee S. 1986. Types of synergy and economic value:
the impact of acquisitions on merging and rival rms.
Strategic Management Journal 7(2): 119 –139.
Chatterjee S, Lubatkin M, Schweiger DM, Weber Y.
1992. Cultural differences and shareholder value in
related mergers: linking equity and human capital.
Strategic Management Journal 13(5): 319 –344.
Clark K, Fujimoto T. 1991. Product Development Perfor-
mance. Harvard Business School Press: Boston, MA.
Cohen WM, Bacdayan P. 1994. Organizational routines
are stored as procedural memory: evidence from a
laboratory study. Organization Science 5(4): 554 –568.
Cohen WM, Burkhart R, Dosi G, Egidi M, Marengo L,
Warglien M, Winter S. 1997. Routines and other
recurring action patterns of organizations: contempo-
rary research issues. Industrial and Corporate Change
5: 653–698.
Cormier S, Hagman J. 1987. Transfer of Learning:
Contemporary Research and Applications. Academic
Press: San Diego, CA.
Datta DK, Grant JH. 1990. Relationships between type of
acquisition, the autonomy given to the acquired rm,
and acquisition success: an empirical analysis. Journal
of Management 16(1): 29 –44.
Dierickx I, Cool K. 1989. Asset stock accumulation and
sustainability of competitive advantage. Management
Science 35(12): 1505–1513.
Dutton JM, Thomas A. 1984. Treating progress functions
as a managerial opportunity. Academy of Management
Review 9: 235–247.
Empson L. 2001. Fear of exploitation and fear of
contamination: impediments to knowledge transfer in
mergers between professional service rms. Human
Relations 54(7): 839–862.
Epple D, Argote L, Devadas R. 1991. Organizational
learning curves: a method for investigating intra-plant
transfer of knowledge acquired through learning by
doing. Organization Science 2: 58 70.
Fowler KL, Schmit DR. 1989. Determinants of tender
offer post-acquisition nancial performance. Strategic
Management Journal 10(4): 339 –350.
Franks J, Harris R, Titman S. 1991. The post-merger
share price performance of acquiring rms. Journal
of Financial Economics 29: 81 –96.
Gavetti G, Levinthal D. 2000. Looking forward and
looking backward: cognitive and experiential search.
Administrative Science Quarterly 45: 113 –137.
Gick ML, Holyoak KJ. 1987. The cognitive basis of
knowledge transfer. In Transfer of Learning: Con-
temporary Research and Applications, Cormier SM,
Hagman JD (eds). Academic Press: New York; 9–47.
Grant RM. 1996. Toward a knowledge-based theory
of the rm. Strategic Management Journal , Winter
Special Issue 17: 109–122.
Haleblian J, Finkelstein S. 1999. The inuence of
organization acquisition experience on acquisition
performance: a behavioral learning theory perspective.
Administrative Science Quarterly 44(1): 29 –56.
Haspeslagh PC, Jemison DB. 1991. Managing Acquisi-
tions. Free Press: New York.
Haveman HA. 1992. Between a rock and a hard place:
organizational change and performance under condi-
tions of fundamental environmental transformation.
Administrative Science Quarterly 37: 48 –75.
Haveman HA. 1993. Organizational size and change:
diversication in the savings and loans industry after
deregulation. Administrative Science Quarterly 38:
Hawawini G, Swary I. 1990. Mergers and Acquisitions in
the U.S. Banking Industry: Evidence from the Capital
Markets. North-Holland: Amsterdam.
Hayward M. 2002. When do rms learn from their
acquisition experience? Evidence from 1990– 1995.
Strategic Management Journal 23(1): 21 –40.
Healy PM, Palepu K, Ruback RS. 1992. Does corporate
performance improve after mergers? Journal of
Financial Economics 31: 135–175.
Henderson RM, Clark KB. 1990. Architectural innova-
tion: the reconguration of existing product technolo-
gies and the failure of established rms. Administrative
Science Quarterly 35: 9 –30.
Jarrell GA, Brickley JA, Netter JM. 1988. The market for
corporate control: the empirical evidence since 1980.
Journal of Economic Perspectives 2: 49–68.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
Deliberate Learning in Corporate Acquisitions 1255
Jemison DB, Sitkin SB. 1986. Corporate acquisitions: a
process perspective. Academy of Management Review
11: 145–163.
Jensen MC, Ruback RS. 1983. The market for corporate
control: the scientic evidence. Journal of Financial
Economics 11: 5–50.
Kale P, Dyer J, Singh H. 2002. Alliance capability, stock
market response, and long-term alliance success: the
role of the alliance function. Strategic Management
Journal 23(8): 747–768.
Kitching J. 1967. Why do mergers miscarry? Harvard
Business Review 45(6): 84 102.
Kogut B, Zander U. 1992. Knowledge of the rm,
combinative capabilities and the replication of
technology. Organization Science 3(3): 383 397.
Krishnan HA, Miller A, Judge WQ. 1997. Diversication
and top management team complementarity: is
performance improved by merging similar or
dissimilar teams? Strategic Management Journal
18(5): 361–374.
Lapre MA, Mukherjee AS, Van Wassenhove LN. 2000.
Behind the learning curve: linking learning activities
to waste reduction. Management Science 46(5):
Levitt B, March JG. 1988. Organizational learning.
Annual Review of Sociology 14: 319 –340.
Lippman S, Rumelt R. 1982. Uncertain imitability: an
analysis of inter-rm differences in efciency under
competition. Bell Journal of Economics 13: (Autumn):
Loderer C, Martin K. 1992. Post-acquisition performance
of acquiring rms. Financial Management 21: 69– 79.
Lubatkin M. 1987. Merger strategies and stockholder
value. Strategic Management Journal 8(1): 39 –53.
Manne HG. 1965. Mergers and the market for corporate
control. Journal of Political Economy 73–74:
March JG, Sproull LS, Tamuz M. 1991. Learning from
samples of one or fewer. Organization Science 2(1):
Marks ML, Mirvis PH. 1985. Merger syndrome: stress
and uncertainty. Mergers and Acquisitions Summer:
Mirvis PH. 1985. Negotiation after the sale: the roots and
ramication of conicts in an acquisition. Journal of
Occupational Behavior 6: 65 –84.
Nelson R, Winter S. 1982. An Evolutionary Theory
of Economic Change. Harvard University Press:
Cambridge, MA.
Neter J, Wasserman W, Kutner MH. 1985. Applied Lin-
ear Statistical Models (2nd edn). Irwin: Homewood,
Nonaka I. 1994. A dynamic theory of knowledge
creation. Organization Science 5(1): 14 37.
Nonaka I, Takeuchi H. 1995. The Knowledge-Creating
Company. Oxford University Press: New York.
Ocasio W. 1997. Towards an attention-based view of the
rm. Strategic Management Journal, Summer Special
Issue 18: 187–206.
Pablo AL. 1994. Determinants of acquisition integration
level: a decision-making perspective. Academy of
Management Journal 37(4): 803 –836.
Pablo AL, Sitkin SB, Jemison DB. 1996. Acquisition
decision-making processes: the central role of risk.
Journal of Management 22(5): 723 –746.
Pennings JM, Barkema H, Douma S. 1994. Organiza-
tional learning and diversication. Academy of Man-
agement Journal 37(3): 608 –640.
Polanyi M. 1962. Personal Knowledge: Toward a Post-
Critical Philosophy. Harper Torchbooks: New York.
Polanyi M. 1966. The Tacit Dimension. Anchor Day
Books: New York.
Puranam P, Singh H, Zollo M. 2004. The coordina-
tion–autonomy tradeoff in technology grafting acqui-
sitions. Working paper, Mack Center for Technological
Innovation, Wharton School.
Rhoades SA. 1994. A summary of merger performance
studies in banking, 1980–93, and assessment of
the ‘operating performance’ and ‘event study’
methodologies. Federal Reserve Bulletin 80(7):
Rogers E. 1980. Diffusion of Innovation. Free Press: New
Rumelt RP. 1974. Strategy, Structure and Economic
Performance. Division of Research, Graduate School
of Business Administration, Harvard University:
Boston, MA.
Rumelt RP. 1984. Towards a strategic theory of the rm.
In Competitive Strategic Management, Lamb RB (ed).
Prentice-Hall: Englewood Cliffs, NJ; 556–570.
Sanchez R, Mahoney JT. 1996. Modularity, exibility
and knowledge management in product and organi-
zation design. Strategic Management Journal, Winter
Special Issue 17: 63–76.
Schweiger DM, Ivancevich JM, Power FR. 1987. Execu-
tive actions for managing human resources before and
after acquisition. Academy of Management Executive
1: 127–138.
Seth A. 1990. Sources of value creation in acquisitions:
an empirical investigation. Strategic Management
Journal 11(6): 431–446.
Shanley MT. 1994. Determinants and consequences
of post-acquisition change. In Managing Corporate
Acquisitions: A Comparative Analysis , Von Krogh G,
Sinatra A, Singh H (eds). Macmillan: London;
Shelton LM. 1988. Strategic business ts and corporate
acquisitions: empirical evidence. Strategic Manage-
ment Journal 9(3): 279 –288.
Shleifer A, Vishny RW. 1994. Takeovers in the 1960s
and the 1980s: evidence and implications. In Funda-
mental Issues in Strategy, Rumelt RP, Schendel DE,
Teece DJ (eds). Harvard Business School Press: Cam-
bridge, MA; 403–422.
Singh H, Montgomery CA. 1987. Corporate acquisition
strategies and economic performance. Strategic
Management Journal 8(4): 377 –386.
Szulanski G. 1997. Exploring internal stickiness: imped-
iments to the transfer of the best practice within the
rm. Strategic Management Journal, Winter Special
Issue 17: 27–44.
Taubman C, Haspeslagh P. 1992. Hanson Plc. INSEAD
Business School case 393-004-1. INSEAD: Fontaine-
bleau, France.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
1256 M. Zollo and H. Singh
Teece DJ, Pisano G, Shuen A. 1997. Dynamic capabili-
ties and strategic management. Strategic Management
Journal 18(7): 509–533.
Thompson JD. 1967. Organizations in Action . McGraw-
Hill: New York.
Walsh JP. 1988. Top management turnover following
acquisitions. Strategic Management Journal 9(2):
Walsh JP, Ellwood JW. 1991. Mergers, acquisitions
and the pruning of managerial deadwood. Strategic
Management Journal 12(3): 201 –217.
Weick K. 1995. Sensemaking in Organizations. Sage:
Thousand Oaks, CA.
Wernerfelt B. 1984. A resource-based view of the rm.
Strategic Management Journal 5(2): 171 –180.
Winter S. 1987. Knowledge and competence as strategic
assets. In The Competitive Challenge: Strategies for
Industrial Innovation and Renewal , Teece DJ (ed).
Ballinger: Cambridge, MA; 159–184.
Winter S. 1995. Four Rs for protability: rents, resources,
routines and replication. In Resource-Based and
Evolutionary Theories of the Firm: Towards a
Synthesis, Montgomery CA (ed). Kluwer: Norwell,
MA; 147–178.
Yelle LE. 1979. The learning curve: historical review
and comprehensive survey. Decision Sciences 10:
Zollo M. 1998. Knowledge codication, process rou-
tinization and the creation of organizational capabili-
ties: post-acquisition management in the U.S. banking
industry. Doctoral dissertation, University of Pennsyl-
Zollo M, Reuer J. 2003. Experience spillovers across
corporate development activities. INSEAD Working
Paper 03–98-SM.
Zollo M, Winter S. 2002. Deliberate learning and
the evolution of dynamic capabilities. Organization
Science 13(3): 339–351.
Copyright 2004 John Wiley & Sons, Ltd. Strat. Mgmt. J.,25: 1233– 1256 (2004)
... In discussing the costs of structural integration theory, it is useful to distinguish between the processes by which structural integration theory is achieved-the set of short-term changes that must be accomplished to create an integrated organization-and the longer-term effects of the final integrated organizational form. The logistics of the transition itself may be costly in terms of time and effort, but these short-term effects are not our focus (Zollo & Singh, 2004). ...
... In discussing the costs of structural integration, it is useful to distinguish between the processes by which structural integration is achieved-the set of short-term changes that must be accomplished to create a structurally integrated organization-and the longer-term effects of the final integrated organizational form. The logistics of the transition from bad structural integration to good structural integrations may be costly in terms of time, resources and other organizational efforts (Zollo & Singh, 2004), but these short-term effects are not our focus. In other words, structural integration when implemented in workplaces could disrupt the target firm's innovative capabilities because it ends its autonomous existence. ...
The objective of this chapter is to reveal the impact of workforce diversity on the innovation level of Lebanese family firms. Data for this research work was collected from different Lebanese sectors during the first quarter of 2021. Based on sample of 647 Lebanese family firms, the results of SEM model show that gender diversity has a positive impact marketing, organizational, and product innovation. In addition, the results indicate that the presence of youth in family firms enhances both marketing and process innovation. Finally, the results do not reveal any impact of gender diversity on process innovation and age diversity on organizational and product innovation.
... In discussing the costs of structural integration theory, it is useful to distinguish between the processes by which structural integration theory is achieved-the set of short-term changes that must be accomplished to create an integrated organization-and the longer-term effects of the final integrated organizational form. The logistics of the transition itself may be costly in terms of time and effort, but these short-term effects are not our focus (Zollo & Singh, 2004). ...
... In discussing the costs of structural integration, it is useful to distinguish between the processes by which structural integration is achieved-the set of short-term changes that must be accomplished to create a structurally integrated organization-and the longer-term effects of the final integrated organizational form. The logistics of the transition from bad structural integration to good structural integrations may be costly in terms of time, resources and other organizational efforts (Zollo & Singh, 2004), but these short-term effects are not our focus. In other words, structural integration when implemented in workplaces could disrupt the target firm's innovative capabilities because it ends its autonomous existence. ...
Despite the international commitment to guarantee the rights of disabled people, including the rights at the workplace, evidence shows that legislations are faced with different barriers to achieving their expected inclusion goals. Even more, the literature reported some counter-effects to the implementation of international and national legislations. Diversity studies have argued that the successful implementation of organizational support strategies is more guaranteed when barriers are well understood. This chapter examines the perceptions of employers and individuals with disabilities, including those with mental disabilities, of the barriers that inhibit diversity management at the workplace. The authors suggest extending the diversity literature by examining other organizational factors that facilitate the employment and sustainability of disabled individuals.
... Researchers have also identified steps to be taken prior to acquisition which can improve the performance post-acquisition. For instance, codifying the knowledge relevant for M&Asuch as development of due diligence manuals and checklists, financial evaluation procedures, and so forth-has strong and positive impact on acquisition performance (Zollo & Singh, 2004). However, prior acquisition experience does not impact post-acquisition performance (King et al., 2004;Zollo & Singh, 2004). ...
... For instance, codifying the knowledge relevant for M&Asuch as development of due diligence manuals and checklists, financial evaluation procedures, and so forth-has strong and positive impact on acquisition performance (Zollo & Singh, 2004). However, prior acquisition experience does not impact post-acquisition performance (King et al., 2004;Zollo & Singh, 2004). But some researchers find that both prior experience of acquisition and codified knowledge about integration after M&A positively influences the integration outcomes (Schriber et al., 2019). ...
Full-text available
Researchers of mergers and acquisitions (M&As) seem to pay less attention to atypical (opportunistic, sudden, and unusual in other ways) acquisitions and subsequent integration processes unfolding in unique contexts. In this study I explore the factors-both similar to and different from what is documented in literature-that contribute to the success of such an atypical and successful acquisition in India. Based on archival data and interviews, I find themes similar to as well as different from what researchers have reported in literature. Unusual focus on clients as well as on talent emerge as distinct features associated with this successful acquisition, but seem underdeveloped in literature. Subsequent researchers may want to examine the veracity and generalizability of these two unique factors' contribution to the success of acquisitions.
... Finally, we were unable to measure the efforts that managers invested in learning and knowledge codification, which may play a role in how experience leads to routine and capability formation (Zollo and Singh, 2004). Moreover, we were unable to directly measure organizational capabilities such as acquisition capabilities or attention modulation capabilities (Castellaneta and Zollo, 2015;Laureiro-Martinez et al., 2019). ...
Momentum theory suggests that acquisition experience leads to acquisition momentum in the form of a higher likelihood of subsequent acquisitions of the same type. However, this argument has been challenged theoretically and empirically. We reconcile conflicting predictions and findings of prior research and extend momentum theory by incorporating activity load as a novel causal mechanism to both replicate the base finding and explain deviations from it. We find that a high activity load due to increased acquisition activity acts as a counterforce to momentum, decreasing the likelihood of subsequent acquisitions of the same type. Moreover, we also find that the interplay of routines, cognitive frames, and activity load causes companies to alternate between different types of acquisitions-from small to large and from large to small-as management engages in attention modulation to preserve momentum. Taken together, our arguments and findings contribute to an improved understanding of temporal patterns of acquisition behavior.
... 6 When there is a chronic shortage of linen, an example of SOPSB is contacting the central service department to increase supplies to avoid the same problem in the future. SOPSB is an intentional learning activity, 7 pushing forward the cause analysis, trying out improvement projects, and promoting continuous improvement. 8 The importance of hospitals learning from their failures hardly needs to be stated. ...
Background and objectives: Nurses are challenged by numerous day-to-day unexpected problems due to poorly performing work systems that hinder patient care. These operational failures persist in hospitals, partly because nurses tend to prefer quick fixes or workarounds over real improvements that prevent recurrence. The aim of this review is to shed light on the barriers to and enablers of nurses' second-order problem-solving behavior and their consequences, so that hospitals can learn from failure and improve organizational outcomes. Methods: We conducted a systematic review, with quantitative, qualitative, and mixed-method articles, searching 6 databases (PubMed, Embase, Web of Science, CINAHL, and Google Scholar) following the Preferred Items for Systematic Review and Meta-Analysis (PRISMA) guidelines. Quality assessment for inclusion was performed by 2 independent authors using the Mixed Methods Appraisal Tool (MMAT). A descriptive synthesis was used for analysis. Results: This study reveals the barriers and enablers for second-order problem-solving behavior, and synthesizes improvement proposals within 3 perspectives, namely the "empowerment" perspective, the "process improvement" perspective, and the "time" perspective. Furthermore, we found that limited attention is given to the patient's perspective, and the existence of a no-action behavior. Conclusion: Although operational failures have several important consequences for hospital staff and organizations, there has been hardly any research into the barriers and enablers that initiate second-order problem-solving behavior; stemming this nursing behavior has thus rarely appeared as a suggestion for improvement.
... When executives change to work at a new firm, in order to alleviate the stimulation and pressure rising from the new organizational environment, they may adjust their behavior accordingly, thereby reducing threats and enhancing their chances of survival. They may also learn from past failures and proactively change their behavior to improve decision accuracy (Zollo and Singh, 2004;Madsen and Desai, 2010;Gong et al., 2019). Consequently, it is an empirical question whether executives' CSR fulfillment will remain relatively stable under the influence of their early experiences and personality traits, or will change in compliance with the new organizational environment. ...
Full-text available
The executives of listed firms play an important role in the fulfillment of corporate social responsibility (CSR). Based on behavioral consistency theory, this study examines the association of CSR performance among multiple firms for the same executive served at different times. By tracking the movement of executives across Chinese listed firms over the period 2010–2019, we find that there is a significantly positive association between the predecessor and the successor firm’s CSR performance for the same executive, implying that an individual’s value and preference for CSR maintain consistency within a certain period of time. We also find that a longer employment gap and lower internal control effectiveness will damage the association of CSR performance between the predecessor and the successor firm. Our results are robust to testing in subsamples and controlling the endogeneity problems. Our conclusion provides a new perspective to understand the influence mechanism of CSR performance in the context of inter-firm executive mobility and provides empirical evidence for listed firms to improve their decision-making in hiring and evaluating executives.
This article applies a mixed‐method approach to explore the complexities of post‐mergers and acquisitions (M&A) integration processes. Extant literature provides significant insights regarding the impact of task and human integration and their influence on post‐integration processes. However, the literature often fails to differentiate between subelements of these two dimensions. This article investigates task and human integration in a cross‐border M&A aimed at efficiency and innovativeness. We highlight the importance of a clear distinction between two subelements of task integration (product harmonization and structural integration) and show that different interorganizational contexts matter. Consequently, we propose a conceptual framework based on two contextual characteristics—source of synergy and choice of location—suggesting that different integration approaches should be applied simultaneously in different contexts of the same post‐M&A organization. Due to the complexity of the post‐M&A integration process, organizing and managing such a process is a challenge to most practitioners. Understanding how to manage different integration dimensions increases the need for prioritization in order to bypass a one‐size‐fits‐all initiative. We suggest that managers should apply a differentiated postintegration approach, depending on specific organizational contexts. A product‐focused approach seems to fit best in a context where target and acquirer remain working in separate locations, but on similar products. A structure‐focused approach is recommended when target and acquirer remain separated, but work on complementary products. Finally, a people‐focused approach is proposed when target and acquirer are colocated.
In view of the growing popularity of diversity, equity, and inclusion (DEI) in academic and policy circles, there is a need for supportive theories. To bridge this gap, this chapter discusses structural integration theory (SIT) as a potent theoretical underpinning for diversity, equity, and inclusion (DEI) discourse. The study made use of secondary resources for critical literature review. In particular, several scholarly articles, texts, and internet resources were reviewed, integrated, and synthesised to provide clarity on the definition of structural integration theory and its assumptions and relevance to discourse on diversity, equity, and inclusion. Ultimately, it was affirmed that structural integration theory is stronger and better for explicating DEI in the workplace and society; hence, it is articulated as a useful theoretical model for devising a multimodal approach to integrating DEI in workplaces.
This study examines whether acquirers support targets' superior corporate social responsibility (CSR) and/or remedy targets' greater CSR problems during the post‐merger integration process using multinomial logit and panel regression models. Furthermore, the study examines the subsequent impact of the acquirers' decision regarding target stakeholder welfare on acquisition performance based on the post‐announcement performance of stock returns. Consequently, we find that acquirers generally do not adopt the superior CSR performance of target firms but instead support their better environmental and product CSR. However, the acquirers do not remedy any of the target's greater CSR problems. This study also reveals that the acquirers supporting targets' superior overall CSR have higher post‐announcement stock returns than others. Meanwhile, post‐announcement stock returns are lower when acquirers fail to fix the targets' greater CSR problems. As this study extends findings from recent scholarship reporting a positive association between CSR and acquisition performance, it sheds light on the importance of stakeholders' welfare in mergers and acquisitions.
Cross-border mergers and acquisitions (M&As) are highly emotional events for the employees of involved organizations. The strength and directionality of emotional reactions can result in positive or negative employee outcomes contributing to success or failure of cross-border M&As. Existing studies on emotions and cross-border M&As have identified various underlying mechanisms and factors that influence employee emotions in cross-border M&A activities, leading to a fragmentation of current research on this topic. In this article, we systematically review the interdisciplinary literature on the role played by emotions in cross-border M&As by analyzing a sample of 78 articles published between 2000 and 2021. We contribute to the current literature by (1) providing a holistic and deeper understanding of the role played by emotions in cross-border M&As; (2) mapping the current state of the interdisciplinary literature on emotions and cross-border M&As; and (3) developing a multi-level framework, and identifying key theories and emerging themes to be examined in future studies.
Full-text available
A sample of 114 mergers made by randomly selected bidders during 1962 to 1983 is classified by changes in the product market opportunities of the bidder firms. Multivariate regression analysis shows that acquisitions which permit the bidder access to new but related markets create the most value.
Full-text available
How do firms create strategically relevant capabilities? In this study, insights from evolutionary economics and organizational learning theories are combined to explore the mechanisms behind the creation of organizational capabilities in the context of infrequent, heterogeneous, and complex administrative tasks. More specifically, the investigation covers the effects of both tacit and codified knowledge accumulation mechanisms on the development of a practice specialized in the management of post-acquisition integration processes. Hypotheses about the performance implications of pre-acquisition resources, post-acquisition integration decisions, and knowledge accumulation and codification processes are tested with primary data collected from a sample of 51 bank holding companies in the United States and Canada, for a total of 577 completed acquisitions. Results show that codification and routinization processes play a key role in shaping the evolution of post-acquisition integration practices, and that both mechanisms have a positive influence on acquisition performance, within specific limitations. The effectiveness of tacit knowledge accumulation is constrained by the degree of homogeneity of past experiences, whereas knowledge codification impacts performance only when high levels of integration are to be achieved. Results also show that greater level of integration have positive implications for acquisition performance and that decisions to replace top management affect performance negatively. Conclusions are drawn about necessary refinements of current theoretical approaches to accommodate complex learning conditions, and the potential implications for the management of acquisitions, as well as other infrequent and complex organizational events, such as strategic alliances and internal reorganizations, are discussed.
Historically, acquisition scholars and practitioners have adopted a choice perspective which portrays the corporate executive analyzing acquisition opportunities as a rational decision maker. This paper suggests that the choice perspective be supplemented with a process perspective which recognizes the acquisition process itself as a potentially important determinant of activities and outcomes. A series of research propositions is offered suggesting how four impediments present in the process itself might affect acquisition outcomes.
A major wave of acquisition activity developed in the US after 1976 (Golbe and White 1988). Evidence on the results of this activity for corporate acquirers is mixed (Porter 1987), while claims about its negative consequences for stakeholders have prompted numerous concerns (Siegfried and Sweeney 1982). The mechanisms by which acquisitions create value are unclear and little attention is devoted to what happens in firms after changes in ownership (Shanley and Correa 1992).1 This chapter studies the determinants and consequences of post-acquisition changes for 51 Fortune 500 acquisitions. Our results suggest that changes are best understood by an incremental management style that avoids disruptions, maintains flexibility, and maximises organisational learning.