Diversification and Performance: Evidence from East Asian Firms

Author(s): Abhirup Chakrabarti, Kulwant Singh and Ishtiaq Mahmood

Source: Strategic Management Journal, Vol. 28, No. 2 (Feb., 2007), pp. 101-120
Published by: Wiley
Stable URL: https://www.jstor.org/stable/20142427
Accessed: 22-11-2023 06:14 +00:00

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Wiley is collaborating with JSTOR to digitize, preserve and extend access to Strategic
Management Journal

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Strategie Management Journal
Strat. Mgmt. J., 28: 101-120 (2007)
Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/smj.572
Received 19 April 2002; Final revision received 3 July 2006

DIVERSIFICATION AND PERFORMANCE: EVIDENCE

FROM EAST ASIAN FIRMS
ABHIRUP CHAKRABARTI,1 KULWANT SINGH2* and ISHTIAQ MAHMOOD2
1 Fuqua School of Business, Duke University, Durham, North Carolina, U.S.A.
2 Department of Business Policy, National University of Singapore, Singapore

We examine the impact of diversification on performance for firms operating in different

institutional environments during a relatively stable period and during a major economy
wide shock. We locate our study in six Asian countries at different levels of institutional
development. Results indicate that diversification negatively impacts performance in more
developed institutional environments while improving performance only in the least developed
environments. Even in the least developed institutional environments, diversification offers limited
benefits when an economy-wide shock strikes. Though successful diversifiers are sometimes
affiliated with business groups, diversification is associated with poorer performance for both
affiliated firms and independent firms. In sum, we find that the outcomes of diversification are
influenced by institutional environments, economic stability and affiliation with business groups.
Copyright ? 2007 John Wiley & Sons, Ltd.

Several studies propose that diversification is more greater degree than firms in relatively developed
likely to be profitable in emerging economies institutional environments.
(Guillen, 2000; Khanna and Palepu, 1997; Kock However, questions remain about the antece
and Guillen, 2001). The underlying argument is dents and consequences of diversified operations
that in emerging economies, intermediate insti by firms in developing economic environments.
tutions?such as financial and market intermedi First, will firms be able to manage their diversi
aries?are inefficient or absent (Alston, Eggerston, fied operations in environments characterized by
and North, 1996; Aoki, 2001; Aron, 2000). In such economy-wide uncertainties? Specifically, would
environments, diversified firms can gain scope firms in less developed institutional environments
and scale advantages from internalizing functions benefit from being diversified when a major eco
typically provided by institutions and markets in nomy-wide shock radically alters the economic
advanced economies. Evidence of extensive and environment? By definition, diversified firms will
apparently profitable diversification in some less be affected by multiple environments simultane
developed institutional environments (Chang and ously undergoing major changes during an eco
Hong, 2002) is consistent with the view that nomy-wide shock, and may suffer from having to
firms in these environments should diversify to a deal with multiple challenges at the same time.
As economic shocks are symptomatic of weak
nesses in emerging economies (Backman, 1999;
Corsetti, Pesenti, and Roubini, 2001), it is impor
Keywords: diversification; firm performance; institutional tant to establish the boundaries or conditionality of
environment; economic shock; Asia
the proposition that diversification produces better
* Correspondence to: Kulwant Singh, Department of Business
Policy, National University of Singapore, 1 Business Link, performance outcomes in less developed institu
Singapore 117592. E-mail: [email protected] tional contexts.

* "*" ?WILEY
* lnt?rSci?nc?@
Copyright ? 2007 John Wiley & Sons, Ltd. %% ?, soMETh,ng^77?

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 102 A. Chakrabarti, K. Singh and I. Mahmood

Second, do group-affiliated firms benefit more an economy-wide shock in less developed institu
or less from diversifying than independent firms? tional environments. We evaluate firm performance
Many legally independent firms in less developed for all analyses but examine the influence of busi
economies are affiliated with business groups, a ness group affiliation.
network form of organization (Khanna and Rivkin, We test these arguments using a sample of 3,117
2001) that enables individual affiliates within a firms operating in Indonesia, Japan, Malaysia, Sin
group to share resources. To the extent that diver gapore, South Korea, and Thailand between 1988
sified groups act as internal markets for affiliated and 2003. These countries play an important role
firms, there might be less need and fewer benefits in the development of East Asia, and are at various
to affiliates diversifying themselves. Yet, access to stages of economic and institutional development.
the group's resources may enhance the outcomes All suffered severe slowdowns during the major
of affiliated firms' diversification. The influence economic shock that struck the region in 1997.
of group affiliation on diversification is especially This allows us to adopt a comparative institutional
intriguing because of the expectation that diver perspective (Guillen, 2001; Hamilton and Biggart,
sification occurs at the group level and not at 1988) to evaluate the diversification-performance
the individual firm level (Chang and Hong, 2002; relationship across different levels of institutional
Khanna and Palepu, 1997; Kock and Guillen, development during and outside an economy-wide
2001), despite anecdotal and empirical evidence shock. In examining the above questions, we seek
suggesting that group-affiliated firms do diversify to clarify the diversification-performance relation
(Chang and Hong, 2002). It is therefore important ship and extend theory and research in Asia by
to establish whether theoretical arguments support adding to the limited empirical strategy research
ing diversification in emerging economies apply to on East Asia and emerging economies (Hoskisson
individual firms affiliated with business groups. et a/., 2000; White, 2002).

In this study, we examine how diversifica

tion affects firm performance in different institu
tional environments, and how an economy-wide THEORY AND HYPOTHESES
shock affects this relationship. Relying on exist
ing arguments, we hypothesize that diversifica The extensive diversification literature documents
tion is more likely to be performance enhanc why firms diversify, the circumstances under which
ing in less developed institutional environments. diversification can improve firm performance, and
Next, we argue that these benefits are predicated the conditions under which firms face increasing
on a relatively stable environment, and hypoth costs of diversification (see Montgomery, 1994;
esize that diversified firms are likely to perform Palich, Cardinal, and Miller, 2000; Ramanujam
worse during an economy-wide shock. This argu and Varadarajan, 1989). Diversification can be
ment rests on the view that firms have difficulty driven by a range of perceived benefits associated
dealing with sudden and major changes (Green with greater market power, more efficient allo
wood and Hinnings, 1996; Hannan and Freeman, cation of resources through internal capital mar
1984; Rajagopalan and Spreitzer, 1997) and that kets, utilization of excess productive factors, more
firms with more complex organizational architec efficient utilization of existing resources in new
tures?which diversified firms embody through settings, or reduced performance variability by
operating in multiple businesses?will experience virtue of a portfolio of imperfectly correlated set
greater difficulty when major changes occur (Han of businesses. However, the benefits of diversifica
nan, Polos, and Carroll, 2003). These difficul tion decline after expansion beyond an optimal or
ties offset the benefits of diversification, such as threshold range, suggesting an 4inverted-U' rela
risk spreading and resource sharing across busi tionship between performance and diversification
nesses, which decline substantially during systemic (Markides, 1992). Greater diversification increases
events such as economy-wide shocks. Finally, we managerial, structural, and organizational com
examine the diversification-performance relation plexity, incurs greater coordination and integration
ship during an economy-wide shock across differ costs, strains top management resources (Grant,
ent institutional contexts, arguing that diversified Jammine, and Thomas, 1988), limits 'organiza
firms suffer greater performance declines during tional attention' (Ocasio, 1997) and inhibits firms'

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 103

ability to respond to major external changes (Don an important determinant of the efficacy of diver
aldson, 2000; Greenwood and Hinnings, 1996; sification (Khanna and Palepu, 1997). Kogut et al.
Hoskisson and Hitt, 1994). Other inefficiencies (2002) attribute differences in diversification pat
and costs arise from conflicting 'dominant logics' terns in five advanced countries (France, Germany,
between businesses, internal capital market con Japan, U.K., and U.S.A.) to nation-specific differ
flicts, and increased control and effort losses due to ences. Matsusaka (1993) shows that in the United
shirking (Markides, 1992). Hence, the difficulties States views toward diversification turned increas
and costs of diversification increase with the scope ingly negative between the 1960s and the 1980s as
of operations, eventually offsetting the benefits of financial institutions developed.
further diversification. Empirical research in devel Khanna and Palepu (1997) argue that greater
oped economies has often concluded that diver diversification may not harm performance in
sification leads to poorer performance, and thus emerging economies because of insufficient mar
carries a 'diversification discount' (Montgomery ket and institutional development. By diversifying,
and Wernerfelt, 1988; Ramanujam and Varadara firms create internal markets that may be more
jan, 1989). In sum, these arguments suggest that effective than inefficient external markets (Ghe
firms can benefit from moderate diversification, mawat and Khanna, 1998; Khanna and Palepu,
but that broader diversification may harm per 1997; Williamson, 1985). These firms enjoy scope
formance (Dess et ai, 1995; Hoskisson and Hitt, and scale advantages from internalizing func
1994; Montgomery and Wernerfelt, 1988; Palich tions provided by external intermediaries or insti
tutions in advanced economies. As intermedi
et al., 2000; Ramanujam and Varadarajan, 1989).
aries are often absent or inefficient in developing
economies, internalization may be viable and prof
Institutional context as a contingency: The itable. Diversified firms do not gain equally in
internal market hypothesis developed institutional environments, as it is more
difficult for internal operations to match relatively
Recent strategy research in emerging economy efficient markets. Similar arguments apply to the
environments (Chang and Hong, 2002; Hoskisson costs of diversification. Efficient markets in devel
et al., 2000; Khanna and Palepu, 1997; Wan and oped economies detect and penalize diversifica
Hoskisson, 2003) has raised interesting questions tion's costs more than the less efficient markets of
about how country differences affect the antece institutionally developing economies. This implies
dents and consequences of firm diversification. In greater net marginal benefits from internalization
this section, we discuss the internal market hypoth in less developed institutional environments.
esis, which proposes that diversification provides Kock and Guillen (2001) propose that protec
greater benefits for firms operating in less institu tionism and other barriers in the less economically
tionally developed environments. and institutionally developed economies distort the
An important implication from existing research value of resources, and make diversification more
is that the gains from diversification often relate to viable than in competitive markets. They propose,
market failure. Market power arguments assume for example, that contacts and connections are
that individual firms cannot contract among them more important than competencies and technologi
selves to match the advantages of the diversi cal abilities for determining the incentives and out
fied firm, while internal capital market arguments comes of diversification in such environments. As
require some failure in information and finan these contacts are not business specific, firms will
cial markets. The resource utilization argument exploit them by diversifying. Some observers (e.g.,
requires that resources be imperfectly tradable, Backman, 1999) note that for many Asian firms
while the reduction of risk proposition requires that diversification is motivated by factors which mar
shareholders value firms' diversification because of ket efficiency arguments do not adequately cap
constraints on their own ability to spread risks. In ture, such as the exploitation of privileged access
general, the potential returns from diversification to information, licenses, and markets. However,
decrease with market and institutional develop these advantages decline with development, sug
ment, so that diversification would not improve gesting that diversification is less beneficial in
firm performance in perfect markets. Therefore, the more developed institutional environments (Kock
degree of market and institutional development is and Guillen, 2001).
Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 104 A. Chakrabarti, K. Singh and I. Mahmood

Therefore, the outcomes of firm diversification more for diversified firms. We expect these effects
will vary across countries, because of the influ to have greater performance impact in less devel
ence of the institutional environment within which oped institutional environments and that business
diversification takes place. We expect that firms group affiliation will play an important role in this
in less institutionally developed economies will issue, but consider these in following sections.
benefit more substantially from diversification than An economy-wide economic shock is a broad,
firms in more institutionally developed economies. major, sudden, and unanticipated change in the
economic environment. An economy-wide shock
Hypothesis 1: The less institutionally developed differs from other hostile environment changes
an economy, the greater the benefits of diversi in two respects. First, economy-wide shocks tend
fication for firm performance. to be systemic?broadly affecting businesses and
industries in a country or region?in contrast to
This section described the internal market hy localized changes that only affect firms in specific
pothesis, that diversification may be more prof industries or technologies. Second, such shocks
itable in less developed economies. However, two tend to be radical, creating discontinuities that dis
issues arise. First, institutional contexts can affect rupt firms, industries, markets and economies, and
both the benefits and the costs of diversification. cause fundamental changes in resource availabil
Research examining the impact of institutional ity, demand conditions, and in the way economic
environments addresses the benefits for diversifica agents operate.
tion, but generally does not consider the manage Economy-wide shocks have two distinct and
ment, organizational, and other costs of diversify sequential effects on firms. First, firms typically
ing in underdeveloped institutional environments. suffer sudden and major disruptions to their busi
nesses, which often lead to declines in revenues
In essence, this perspective views weak institu
tional environments as increasing the marginal and profits. These primarily result from disruptions
benefits of diversification but not marginal costs. in demand, markets, suppliers, and buyers; rapid
We return to this issue in our discussion below. increases in financial costs and risks; and changes
The second issue relates to an assumption under in exchange rates (Singh and Yip, 2000). Uncer
lying the internal market hypothesis, that firms can tainty about the depth, duration, and consequences
diversify non-systematic risk by moving resources of the shock often compound these performance
across unrelated units. It is less clear, however, declines. Second, and usually after the causes,
how diversification affects performance when risks extent, and consequences of the shock become
are correlated, as in the case of a major economy clearer, firms restructure to adapt to the changed
wide shock. Research has not evaluated whether environment. In this paper, we focus on the first

the performance benefits of diversification in times effect, the impact of the shock on firms before they
of relative stability transfer to when firms face adapt.
major and broad external changes to the economy. The spreading of risks benefit of diversification
In the following sections, we discuss the impact may not apply during an economy-wide shock.
of an economy-wide shock on firms to clarify the Lubatkin and Chatterjee (1994) argue that the port

boundary conditions of the internal market hypoth

folio effect for securities is not directly appli
esis. We then examine how the effects of diver cable to firm diversification, suggesting limited
sification on performance may vary with group risk-spreading benefits when economic conditions
affiliation. change broadly. The advantages that diversified
firms enjoy in less developed institutional envi
ronments will also decline, as privileged access to
Economic stability as a contingency:
resources and internal transfers provide relatively
Economy-wide shock
few benefits during periods of sudden, widespread,
In this section, we argue that greater diversifica and drastic economic slowdown. When many or all
tion is associated with poorer performance during of a diversified firm's markets are affected during
economy-wide shocks. We rely on two main argu an economy-wide shock, it will be less able to shift
ments, that an economy-wide shock (1) reduces resources from strong to weaker business. Some
the internal market benefits of diversification, and businesses that were viable because of resource
(2) increases management and organizational costs transfers will suffer in their absence or reduction.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. /., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 105

An organization's architecture is determined by and Hinnings, 1996; Hannan and Freeman, 1984;
its formal structure, processes, rules, and coordi Newman, 2000; Rajagopalan and Spreitzer, 1997).
nating and governance mechanisms (Hannan et al., Consequently, in addition to reducing the ben
2003), with greater variety in these increasing efits of diversification, an economy-wide shock
organizations' architectural intricacy or complex increases the costs of diversification. An economy
ity. Diversified firms embody greater architectural wide shock decreases the benefits of diversification
complexity through the creation of separate struc by diluting the benefits of internal markets and
tures, processes, and rules for each operation and increases the costs of diversification by making
business, and through the use of structural mech it more difficult to manage multiple challenges
anisms for coordination and governance. Hence at the same time. Collectively, the fall in bene
diversified firms embody greater architectural com fits and the rise in costs of diversification imply
plexity to have the requisite variety to deal with that broader exposure to simultaneous, sudden,
the multiple environments they operate in (Ashby, and widespread adverse changes in their envi
1969). This complexity reduces their flexibility and ronments will cause diversified firms to suffer
responsiveness to external change. Structural solu greater performance declines during an economy
tions, such as the adoption of the M-form structure, wide shock.
may moderate but will not eliminate architectural
complexity unless firms allow divisions to operate
Hypothesis 2: The more diversified a firm, the
independently; however, this prevents coordina
greater the decline in its performance during an
tion and resource sharing across divisions, limit
economy-wide shock.
ing the benefits of diversification (Markides and
Williamson, 1996).
The cascading organizational change framework Interface between economy-wide shock and
introduced by Hannan et al. (2003) illustrates the institutional contexts
scale and scope of challenges diversified firms
face during an economy-wide shock. During an The change in firm performance during an eco
economy-wide shock, actions by individual busi nomy-wide shock will be influenced by inter
nesses or units lead to a cascade of reactions in nal factors such as diversification as discussed
other units and businesses within the firm, with above, and by external factors such as the state
a greater cascade likely in firms with greater of institutional development, the subject of our
architectural complexity. Consequently, diversi next hypothesis. In general, the difficulties that
fied firms are likely to face greater cascades diversified firms face during a major economic
of strategic, organizational, resource, information, shock are likely to be greater in less developed
institutional environments. There has been little
and management challenges than more focused
firms during an economy-wide shock. These cas empirical research on how the interface of firm
diversification and institutional environments influ
cades of changes strain and divert managerial
attention, and hinder appropriate response (Hannan ences firm performance during an economy-wide
et al, 2003; Ocasio, 1997). shock.
The trend for diversified firms to deal with In less developed institutional environments,
greater architectural complexity by introducing institutions are likely to be less effective in mobi
rigid financial controls and routines further lizing and provisioning resources to firms during
reduces their flexibility and responsiveness to an economy-wide shock. The relatively unsophis
external change (Greenwood and Hinnings, 1996; ticated financial and market institutions in less
Hoskisson and Hitt, 1994). Donaldson (2000) institutionally developed economies are likely to
argues that diversified firms generally exhibit be less effective at providing the liquidity firms
greater misfit between structures and environments typically need during an economy-wide shock.
because of wider spans of control and the greater The relative weakness of regulations and regula
supervisory burden of CEOs, and therefore are tory agencies, and of professional and technical
less able to change. Studies in the organizational intermediaries, may result in organizations hav
change, restructuring, and ecology literatures ing less information about the causes and impact
further indicate that firms have difficulty dealing of the shock, and less assistance in understand
with sudden, broad, and major change (Greenwood ing and dealing with the consequences of such
Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
106 A. Chakrabarti, K. Singh and I. Mahmood

shocks. This will hurt the performance of diver firms moderates problems associated with indi
sified firms more because of the greater scope vidual firm diversification, such as organizational
of the challenges they face during an economy architecture complexity, conflicting 'dominant log
wide shock. In essence, the institutional weak ics,' and loss of focus.
nesses that permit firms to benefit more substan This raises two issues related to groups and
tially from diversifying during relatively stable diversification that have not been adequately ad
periods will cause these firms to suffer greater dressed. First, why would group-affiliated firms
declines in performance when an economy-wide diversify when their groups can do so more effec
shock strikes. tively?1 If business groups are efficient mecha
The arguments leading to Hypothesis 2 suggest nisms for diversifying in less developed institu
an economy-wide shock will reduce the benefits tional environments, groups would diversify
and increase the costs of diversification. Here, we through the creation of firms for each business,
expect the decline in the benefits and the increase reducing the incentive for these affiliated firms
in the costs of diversification during an economy to diversify. Yet, group-affiliated firms differ in
wide shock to be greater in less developed insti their scale, scope, and profitability of operations,
tutional contexts. Thus, we expect the net per suggesting varying levels and success of diversi
formance benefits that diversified firms enjoy in fication (Chang and Hong, 2002). The business
relatively stable environments to decline during a group literature has paid little attention to group
shock, and to decline more in less developed insti affiliated firms' diversification, perhaps implicitly
tutional environments. relying on the proposition that group diversifica
tion substitutes for and precludes affiliated-firm
diversification.
Hypothesis 3: The less developed the institu
Second, what are the outcomes of group
tional environment, the greater the decline in
affiliated firms' diversification? Group-affiliated
performance of diversified firms during an
firms may be able to diversify successfully by
economy-wide shock.
exploiting spillovers from group-level resource
transfers. They may also be able to mobilize
The impact of business group affiliation resources more readily or at lower cost from
external agencies, because group affiliation may
Business groups are an important feature in many provide reputation benefits and privileged access.
less developed institutional environments (see Groups are also likely to develop greater
Chang and Hong, 2002; Khanna and Rivkin, managerial and organizational sophistication and
2001). Business group affiliation may significantly resources from managing their more complex
affect the outcomes of firm-level diversification structures and diverse operations. These may spill
and the manner in which an economy-wide shock over into improved diversification competencies at
affects group-affiliated firms because these firms the affiliated-firm level. Collectively, these factors
have access to network-level resources. may improve the outcomes of diversification for
Business groups?networks of legally indepen affiliated firms. However, group objectives may
dent firms linked by a set of formal and infor restrict affiliated firms' diversification, limiting
mal ties that coordinate their actions?typically performance gains. For example, affiliated firms'
attempt to overcome or exploit inefficient or absent
markets and institutions in emerging economies 1 There are several reasons for doing so, all of which relate to
by diversifying broadly through the creation of imperfect substitutability of group diversification for affiliated
firm diversification. These include the advantages of diversifying
networks of related firms (Biggart and Hamil
on the basis of resources or competencies that are available in an
ton, 1992; Khanna and Palepu, 1999; Kock and affiliated firm, but are not easily transferable or decomposable,
Guillen, 2001). Business groups comprise firms even to related organizations (Siggelkow and Levinthal, 2003);
the desire to reduce variance in the affiliated firm's performance,
operating in different industries, whose activities
which may be an important consideration for firms with concen
are coordinated by a central firm or organization. trated ownership (Thomsen and Pedersen, 2000); the possibility
Coordination centers on, but is not restricted to, that market opportunities may be too small to justify the estab
lishment of a separate unit within the group; and agency-related
resource transfers between affiliated firms, with
issues at the firm level. We do not explore these motives in
resources moved to where they have the greatest this paper, but investigate their manifestation in the outcomes of
affiliated firms' diversification.
effect. The establishment of separate but affiliated

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. 7., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 107

diversification may be hindered by the need firms' diversification, and whether such affiliation
to contribute resources to the group (Lincoln, buffers the impact of an economy-wide shock in
Gerlach, and Ahmadjian, 1996), or by constraints different institutional environments.
on expanding into businesses where other
affiliated firms are incumbents or prospective
entrants. It is an empirical issue whether group DATA AND VARIABLES
affiliation provides net positive benefits for firm
diversification. We examine this as a research We locate our study in six East Asia coun
question. tries?Indonesia, Japan, Malaysia, Singapore,
We next evaluate whether group affiliation South Korea, and Thailand?for the 1988-2003
buffers the negative impact of an economy-wide period. These countries are important economies
shock. Spillover benefits from resource sharing in the region, are at various stages of economic
within groups are likely to decline substantially and institutional development, and suffered broad
during an economy-wide shock, as most or all and severe slowdowns during the major economic
affiliated firms will be affected simultaneously. shock that struck the region in 1997. Our empirical
Firms that relied on internal resource transfers context provides a conservative test of the benefits
to support their diversification will be affected of diversification, as very rapid economic growth
more severely if such transfers decline because over the two decades prior to 1997 provided a
of resource shortages. Firms less affected by an munificent environment that allowed even rela
economy-wide shock may be required to share tively inefficient firms to prosper (Corsetti, Pesenti,
resources with other affiliated firms, harming the and Roubini, 1999).
outcomes of their diversification. Other group We selected 19 manufacturing industries gen
policies may restrict firm choices and flexibil erally represented in our sample countries using
ity. Negative reputation effects from poorer group all firms listed in the Osiris database (Compact
performance may cause affiliated firms to lose CD-ROM, Disclosure, Inc.). Restricting the study
the privileged access they often enjoy to exter to an industrial sample controls for technological
nal resources as a result of their group affilia influences on diversification and allows better iso
tion. However, the converse is also possible, that lation of the relationship between diversification,
an affiliated firm may enjoy superior access to performance and institutions. We included firms
external resources during an economy-wide shock that entered our sample after our start year of
because of its group affiliation. Meanwhile, diver 1988 and retained them until they exited from our
sified independent firms face similar challenges of sample. The final sample comprised 3,117 firms.
broad exposure to an economy-wide shock, though Of these, 942 were single-business firms at some
without the burden or benefits of group affilia point during the study period. We obtained macro
tion. economic and institutional data from Economist
In view of these conflicting effects, we exam Intelligence Unit's CountryData and Country Indi
ine the net benefits of group affiliation during an cators databases.
economy-wide shock as an empirical issue. We
also examine these effects across institutional envi
Measures
ronments, as institutional and business group dif
ferences may affect outcomes. Diversification
In summary, we propose that individual firms
We measured diversification with the widely used
operating in less developed institutional environ
ments benefit more from diversification than firms entropy measure (Jacquemin and Berry, 1979;
Palepu, 1985), calculated as follows:2
operating in more developed institutional envi
ronments (Hypothesis 1). However, these benefits
are limited to relatively stable periods, so that
N1
diversified firms will suffer greater performance
declines during an economy-wide shock (Hypoth
esis 2), especially if they operate in less developed
2 We did not adopt the refinement of the entropy measure
environments (Hypothesis 3). Finally, we evaluate proposed by (Raghunathan, 1995) as the disaggregated data
how business group affiliation influences affiliated required were not available.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 108 A. Chakrabarti, K. Singh and I. Mahmood

where Pt is the share of the /th segment? development across countries, general conformity
operationalized in Osiris at the four-digit SIC among institutional measures, and equivalence
level?of the firm's total sales. We used the between these measures and per capita GNI. One
quadratic measure of entropy to test for nonlinear exception is that despite Japan having the high
relationships. As an alternative measure of diver est per capita GNI, all measures ranked it below
sification, the Herfindahl Index, was highly corre Singapore in institutional development.
lated with entropy, we only report results using the
entropy measure.
Economy-wide shock
Most countries in South East and East Asia were
Performance
struck by their most severe economic shock in
We used return on assets (ROA) with a 1-year lag decades in 1997 and 1998 (Corsetti et ai, 1999;
as the measure of firm performance for the test of Dornbusch, Park, and Ciaessens, 2000). The sud
Hypothesis 1. For Hypotheses 2 and 3, we use log denness and severity of the shock were unantic
of next period ROA as the measure of change in ipated and caused countries to suffer rapid and
performance. We discuss this measure below. significant adverse movements in exchange rates,
foreign investment, and trade, leading to major
Institutional environment economic recessions. Foreign investors pulled out;
access to funds was curtailed, many firms defaul
Despite widespread acceptance of the concept and ted, shut down or reduced their operations, and
importance of institutions, there are no theoret many jobs were lost. The shock was a major cri
ically established empirical measures of institu sis for all sample countries. Average ROA for our
tions (Aron, 2000). We therefore adopt a compar sample declined between 1995 and 1998, averag
ative approach to test the impact of institutional ing 0.074, 0.071, 0.057, and 0.039, respectively.
environments, evaluating hypotheses by estimating As we use a short window for this analysis, our
country-level models. Differences across countries study design partially controls for the endogenous
would be indicative of institutional differences. effects of firm adaptation and external institutional
Table 1 provides information on institutional change, allowing us to focus on the impact of the
environments across countries, summarizing three shock on firm performance.
different measures of institutional development We used three complementary measures of the
and one measure of economic development, per shock. The first is industry market capitalization,
capita gross national income (GNI). This table which we computed by summing the market cap
indicates substantial differences in institutional italization of all firms in our dataset in each

Table 1. Measures and rankings of institutional environments

EIU indicator3 Euromoney country Composite ICRG Per capita

creditworthinessb risk rating0 GNId

Singapore 8.81 (1) 93.7(1) 92.5 (1) 20,690(2)

Japan 7.39 (2) 93 (2) 87.3 (2) 34,010(1)
S. Korea 7.12 (3) 80.3 (3) 79.3 (3) 9,930(3)
Malaysia 7.06 (4) 79.4 (4) 76.3 (4) 3,540(4)
Thailand 6.78 (5) 64.6 (6) 71.3 (5) 2,000(5)
Indonesia 5.89 (6) 68.9 (5) 66.3 (6) 710(6)

a EIU Country Indicators database. Computed from 37 items rating various aspects of the political environment, the macro-economic
environment, market opportunities, policies relating to free enterprise, competition and foreign investment, foreign trade and exchange
controls, taxes, financing, labor markets, and infrastructure. Data presented are for 2003.
b Assesses country risk based on nine weighted categories that include debt, economic performance, political risk, and access to
financial and capital markets. Based on polls of economists and political analysts. Measured for 1997. Reported in World Development
Indicators (1998).
c Information on 22 components is grouped into three major risk classes: political, financial, and economic. These are transformed into
a single numerical risk measure. A score below 50 implies very high risk, and one above 80 implies very low risk based. Measured
for 1997. Based on PRS Group monthly updates, and reported in World Development Indicators (1998).
d Real per capita GNI for 2003, in thousands of U.S. dollars per year, obtained from Asian Development Bank.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 109

industry, within each country, for each year of other business groups, to account for substantial
analysis. We used a similar approach to measure differences in strategies, resources, and perfor
industry average ROA, our second measure of the mance.
shock. Industry market capitalization and industry
average ROA control for differences in oppor
Control variables
tunities and the impact of the shock. The third
measure was per capita GNI, which is a broad We included several control variables. Firm size,
indicator of the economic environment. Collec based on total assets (millions of U.S. dollars,
tively, these measures comprehensively indicate logged), controlled for size-related impact on per
the economic environment firms operated in before formance and for the greater tolerance that large
and during the shock. These measures also con firms may have to economy-wide shocks. Age
trol for variation in the economy-wide shock in (in years), current ratio (current liabilities/total
each country, as the shock affected firms most liabilities), and debt ratio (debt/equity) measured
directly through changes in industry-level opportu resource availability and constraints for each firm,
nities and profitability. For the test of Hypotheses and indirectly controlled for affiliated firms' use of
2 and 3, we use 1-year changes in these val group resources (Chang and Hong, 2002). Finally,
ues to detect the impact of the economy-wide an indicator dummy variable in our test of Hypoth
shock.
esis 1 distinguished between the period prior to
the economy-wide shock (1988-96; pre-shock =
1) and after it (1999-2003; pre-shock = 0). Con
Business groups
cerns about time period effects were reduced by
We used several different sources to complement findings of marginal time effects by Khanna and
the Osiris database to identify business group Rivkin (2001), whose study included several of
affiliation, relying primarily on published papers, our sample countries for overlapping time peri
researchers, industry and market publications, and ods.
firm websites. For Japanese business groups, we Tables 2 and 3 present descriptive statistics.
also distinguished between horizontal and vertical Table 3 shows that group affiliation was only
business groups, to account for potential differ associated with lower diversification in a few
ences in strategies (Lincoln et al., 1996). Hori cases, confirming the value of testing diversifi
zontal business groups tend to diversify broadly cation's outcomes for affiliated firms. Affiliated
across industries, which contrasts with vertical firms achieved lower ROA than independent firms
business groups' focus on forward or backward in most countries for both non-shock and shock
integration within the same industry. Following periods, a pattern consistent with earlier research.
Chang and Hong (2002), we distinguish between Khanna and Rivkin (2001) show that independent
the 30 largest business groups in Korea and firms achieved higher average ROA than groups

Table 2. Descriptive statistics

Variables AT Mean S.D. 1

ROA 47,425 0.059 0.217 1.000

Entropy 41,609 0.521 0.466 -0.071 1.000
Size 48,947 11.755 2.116 -0.050 0.177 1.000
Debt ratio 49,033 0.596 10.963 -0.040 0.034 0.035 1.000
Credit ratio 48,924 6.455 0.002 -0.025 -0.037 0.094
1.971 1.000
Age 49,033 35.430 23.355 -0.125 0.241 0.555 -0.007 -0.033 1.000
Industry ROA 49,033 0.059 0.056 0.255 -0.159 -0.216 -0.009 0.004 -0.305 1.000
Industry market 49,026 0.575 3.414 -0.017 -0.047 -0.041 -0.014 0.003 0.056 -0.061 1.000
capitalization
GNI per capita 48,985 20.464 13.893 -0.096 0.294 0.473 0.002 0.007 0.655 -0.371 0.103

All values are in thousands of U.S. dollars (except age, debt ratio, current ratio).
All values >0.01 or >0.01 are significant at the 0.05 level or lower.
Number of firms = 3117.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 110 A. Chakrabarti, K. Singh and I. Mahmood

O O oo ? CN
oo in
Ooo^r
en or-eno rf CN
CN en
Z feo
G on
O
-o
CN
C or?>oen
in r ^ cn ^>
m in
en en O (N
^?- (o CN
m in
en en
O ?o OO O ?
. Oh
oC o3 o ^h en
i> on
O ON
On rt
? en
r- in
en (N
C Z 00
O CN
?o ? ? ? ? oo?
O CN CN

H
'cd
^en
qcN?;r
es O^: oo
^r m
"3 ^ in
ON

o
VO 00
O 00 (N on
O
en
do do
^ CN O (N O ??
oo o
I Oh in ^o ^1" 00 O -H
oo on o
O CN
O o do*
en rt
CN en or- oo? CN Tt
CN en
Z feo dowin
o ow **
cd
Oh
;r-CN
<-< CN
o ^t ?n *?
O CN
^doo ^^
en
^^O^t^h oeno^r <^
O
do do
O ? en Tt
ON

oi
en ^r
do

C S3
o o
I Oh
o CN ^r
oo on
O 1-H
r- o
in tj
en cN
O~
CN CN
en ^r
oo
-?t in Tf
in
Z ?b OO
*-< ex
CN o in * ^t- ?< ?
v. ^ ^ oo en
^ O ^H
o 6O '?T O ^h . m Tj- oo q - s ^. ^.
?4 O &
CN
O O S O O r^ O O q< o o

00 ^D oo ^o
ao in Tj
O ?'
oo oo
Oen vo
? in \o
in ^i
Oo do Oo do
ON in Tf
e2 a

S2
in ^t CN r? in on
r- in
O CN
oo OO oo
Ti"
00

in in oo r^
3% ^en o
OO On
^t CN CN CN en ^ *
&' OO
on ' ? ?" ?oo
CN

S?

t\ ^O CN ^t o
On -h as
Ooo*-<o o o
CN oo
o^h ON
o cN r^
oo OO
S?

"2 in o inoino ^oo

o^h in
U Oenm^
l Oh
rf oo
OO en en cN en en
.o o *
2 w6
CN
ON
O
en

Oh CN oo r^ t^ en
O B2 enin en^f in O ^o in
I

en ^t
O on o o O o
oo
?-H

On

S "^3

on
on

ss
lis c?5 tq
c00

Copyright ? 2007 John Wiley & Sons, Ltd.

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 111

in 8 of 14 countries, and that group affiliation had in diversification during 1997 and 1998, making a
positive impact on performance in only three coun test of the impact of diversification on performance
tries.
with this estimator difficult. The simplest approach
to measure change in performance?using change
in ROA as the dependent variable?is subject
METHODS to problems of reliability associated with change
scores (Bergh and Fairbank, 2002). We therefore
To test Hypothesis 1, we used a panel design on use the log transformation of next period ROA as
observations running from 1988 to 2003, for each the dependent variable.4 We introduced 17 industry
country, excluding the shock years of 1997 and indicator variables for OLS regression models.
1998. The Hausman test (Hausman, 1978) indi
cated that omitted firm-level unobserved effects
correlated with the regressors for all samples RESULTS
except for Thailand. This ruled out pooled OLS
and random effects estimation because these mod Hypothesis 1: Institutional context
els include omitted variables in the error term and
Table 4 presents results of our fixed-effects models
assume that these omitted variables are not arbi
for the test of Hypothesis 1. To facilitate discus
trarily correlated with the regressors (Wooldridge, sion, countries are listed in order of institutional
2002). We therefore estimated fixed-effect mod development (see Table 1) from the left.
els using country samples. We tested the results While the diversification was negatively asso
for Thailand using fixed and random effects mod
ciated with performance for the full sample, this
els, and found that though coefficients varied, the association differed across countries. Diversifica
direction and significance of all variables were tion was positively associated with ROA for firms
consistent. We therefore report fixed-effects mod in Indonesia, the country with the least developed
els for Thailand. We further split the data into institutional environment in our sample. In con
group and independent samples, to test group affil trast, diversification was negatively associated with
iation effects.
performance for firms in Japan and Korea, coun
We tested whether the data for the 1988-2003
tries with relatively developed institutional envi
period (less the shock years) were poolable by ronments. These results support Hypothesis 1. The
running tests of stability for 3-year periods for absence of a diversification-performance associ
each country. Our first test compared results for ation for firms in Malaysia and Thailand, both
the 1988-90 period with those for the 1991-93 economies with relatively weak institutions, sug
period. We then compared results for the 1988-93 gests that the internal market hypothesis only holds
period with those for 1994-96, repeating this pro in substantially underdeveloped institutional envi
cess for the entire period. Chow tests for stabil ronments. Hypothesis 1, however, is not supported
ity showed that data pooling posed no problems, by firms operating in Singapore. We discuss the
except for some periods for Japan.3 However, we implications of this result below.
believe that this does not pose a significant prob We tested the sensitivity of results to the time
lem, as our pooling of observations was driven by period, dropping the boundary years of 1996 and
concerns of firm heterogeneity, not by concerns of 1999 from the analyses reported in Table 4. We
stability over time. also introduced variables to measure the propor
To test whether diversified firms performed tion of foreign sales for each firm, to control for
worse during an economy-wide shock (Hypotheses the possibility that foreign operations may moder
2 and 3), we estimated OLS models using obser ate the outcomes of diversification. This analysis
vations for the shock years of 1997 and 1998. We was constrained by inconsistent reporting of for
were unable to test the impact of diversification eign sales, so that data were only available for
on performance using the first difference model approximately 33 percent of our sample. Results
as many firms did not report substantial changes remained substantially unchanged.

3 We do not present the results of the Chow test because of space 4 Suppose In Y = a + bX. Differentiating, we get (l/Y)?Y =
constraints. We thank an anonymous reviewer for suggesting the b?X, i.e., b = (?Y/Y)/?X. This is the proportional change in Y
pooling test. for an infinitesimally small change in X.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 112 A. Chakrabarti, K. Singh and I. Mahmood

Table 4. Effects of diversification on ROA in periods of relative stability

Variable Full Singapore Japan Korea Malaysia Thailand Indonesia

sample

Entropy -0.018*** -0.012 -0.024*** -0.017* -0.017 -0.003 0.110***

(0.005) (0.016) (0.007) (0.009) (0.013) (0.023) (0.038)
Firm size (log) 0.011*** 0.023*" 0.027*** 0.002 0.001 0.007 0.063***
(0.002) (0.005) (0.003) (0.003) (0.004) (0.005) (0.008)
Debt ratio -0.001*** 0.000 -0.001*** -0.000 -0.002 -0.002*** -0.001**
(0.000) (0.000) (0.000) (0.000) (0.001) (0.000) (0.001)
Current ratio 0.001 -0.000 -0.006*** 0.002*** 0.004*** 0.004*** 0.007***
(0.000) (0.000) (0.001) (0.001) (0.001) (0.000) (0.001)
Firm age -0.002*** -0.004 0.001 -0.001 -0.004 0.000 -0.009
(0.001) (0.003) (0.001) (0.002) (0.003) (0.004) (0.006)
Industry ROA 0.856*** 0.789*" 0.535*** 0.926*** 0.853*** 0.901*** 0.649***
(0.028) (0.062) (0.084) (0.068) (0.079) (0.075) (0.066)
Industry market 0.003*** 0.024 -0.001 0.001 -0.054 0.001 -0.005
capitalization (0.001) (0.022) (0.001) (0.001) (0.043) (0.069) (0.008)
GNI per capita -0.001** -0.001 0.000 -0.001 -0.011 0.002 0.009
(0.000) (0.002) (0.001) (0.002) (0.013) (0.022) (0.088)
Pre-shock 0.000 0.009 0.005 0.005 0.023 -0.007 -0.030
dummy (0.005) (0.019) (0.010) (0.012) (0.021) (0.040) (0.063)
Constant 0.986*** 0.848*" 0.621*** 1.017*** 1.127*** 0.920*** 0.507***
(0.025) (0.078) (0.074) (0.039) (0.080) (0.093) (0.167)
N 34,938 2,261 14,770 10,658 3,905 1,836 1,536
F 174.54*** 45.03*** 30.03*** 30.32*** 86.76*** 32.70*** 34.53***
Adjusted R2 0.071 0.194 0.036 0.043 0.214 0.152 0.198

Dependent variable is firm ROA.

Industry market capitalization and GNI per capita scaled to $/1000.
Data are for 1988-2003, excluding shock years of 1997 and 1998.
Fixed effects regressions with robust standard errors in parentheses; p <0.01; ** p <0.05; * p < 0.1

Group affiliation diversification, consistent with the prediction that

diversification has negative impact in more devel
We next consider how business group affiliation oped institutional environments. This suggests that
affects the diversification-performance relation diversification by firms affiliated with horizontal
ship. Table 5 replicates Table 4, but splits coun groups in Japan compounds the error of these
try samples into group-affiliated and independent groups' broad diversification (Lincoln et al., 1996).
firms, for the 1988-2003 period, excluding shock Diversification also hurt the performance of firms
years of 1997 and 1998. Results show that group affiliated with vertical groups in Japan, possibly
affiliation differentiated the outcomes of diversi because these groups' focus on diversification in
fication in four countries. Group affiliation was vertically integrated businesses did not provide
most helpful in Singapore, where diversification spillover knowledge or resources to help the hori
had a positive impact on performance for affiliated zontal diversification of their affiliated firms. Affil
firms and a negative impact for independent firms. iation with top 30 and the other business groups
Group affiliation also had a positive impact on in Korea had negative outcomes, possibly because
diversification outcomes in Thailand, where diver business groups in Korea have relatively high lev
sification had no performance impact for indepen els of horizontal diversification. Group affiliation
dent firms. As these countries differ substantially was not relevant in Malaysia, where diversification
in institutional development, it is likely that other did not impact ROA, or in Indonesia, where all
country-specific factors make diversification more firms gained from diversifying. Independent firms
beneficial for group-affiliated firms here than in did not gain from diversifying in all countries,
other countries. except for Indonesia.
Firms affiliated with business groups in Japan The results of Tables 4 and 5 indicate that diver
suffered poorer performance with greater sification tends to improve performance in less
Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. /., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 113

r^rnio^H(-ocNoo(NinON^tr^iO'-^vsOvD(NOeNOP^;? ^
^o^ooor-^oQOO-HOooooo-^meNir-ONO>-^ipr-m
? OOOOOOOOOvOOOOO-HOOroeN^^-CM
dodododododododododo ^t c5
0^o>cN^^^-^<r^o>^^uo^^^r~-ooor--'^-eNf^:* *?i
O oooooooooovo^ooo^oomcN^r-^
OOrtrtT-HQOQOOQr^OO^^C^QOOeM^^?Nr
dddddddddddddddddddd en o
WW|WW|WW|W|www_

oo m(0-H\0(NOmMir)VDninmoioOh^,?-HM* m
Ttm^oOQ^OQOcMONr-^oom<N?ort^-Hcn^r
ooooooooooooo^oooor-^H^oo?<
^ b?
0000000000?^OOOOOOOO
! w w i w w w w i w i w w (NO
w^
CN m en cN o r-< (N oor--env>ovo<NeNooooeNenT~l* r
ooooooo ?lOOCNOincNmONinenvo^OACN
OO^?-^-HOOOOO^t^^en?N
O dddddddddddddddddd??o
WW|WW|wwww,ww

oc po oomenrtencNr-^cxDTtTfvoooONininvo^DcMO* >sO
8rtQ50QOOOO'?OMOO'H-H(NTtMOO\(N
!WW 000000000000000000^0 ~ o
Z &b
c | W W | ^ W | W | W W w ^
aG
73
VO Tf O < NrHtn^xin(N(NO0\O^?>O^r<i2P? oo
^HCN^?OQQOQO<nO"?l-^D'^'-'T:tm(N000>O,!t
O0000000?5or-^OOOOOOON^?OCM(N
dddddddddddddddddddd en o
I I I
CN

T3
C I Oh
e =?
oo
:888; '800008800
> cn en vo *-h cn <?< en en ^o r^ ^ *?t :
^O Tt Tt o ?n
OOOOOOOOOO^OOOOOOOOO
W W I W W I W W W W|W w
cd

S o m\DO\h-HQhro^c<)(Ni-TH^fn^oO,t'vf(N* O
3 O & 83888888888288888838 ?-P3
Oh
OS-H
O?
dodddooododddddddd-Hd
I W I W I W W W W W|W W w
Ko
^^^^^H^i^^oo^^O^CNMinOio^OcnO* en
cn^ir^OOQ^dQdcNr^OOQO^cN^tH'-Hcnun
Oh O
OOOOOOOOOOO^OOOOOO-H-^^eo-H
dddddddddddddddddddd oo o
cd Iw1ww^ww1w1www

oo ^^?\Oi-'HOh^^(S(NmO'HO^MiOO00iO s en j
r-4-H(DOOOQOQQ,OeNQOQO^'-H^DONenr-?o
Z o?
???????????????????? do
OOOOOOOOOO^-HOOOOOOinOo^^tO
I I I
w en V
Ci.

-O
O CN en on ^o CN T-H 10 '?i en O ^D en CN 1 ?< en rf ON Tt Tf
?2 oooSSSSb?o?^oo?oo?o^^lt^o
???????????????????? ho o as u
I w 1 w 1 w 1 w w w 1. U H Ui

<
w fl ta
G Oh -o r^ ?"
O O S H0Mn^OO(NnNNfn^H(N)HHH00rHOh. . . J? <* c-
s en 3
NO en?(NQQQOOQQr-^OQOOQ^(NioOTj-(N (5 ON
c*
oc
d0000000000o?000000o
0OO000OOOO,0'-H0O0O0Own?' ^jTKo
00 O
?og

Si.=3^ OHT?mmTi-HoiO' oo(NenoN(NcNHH?N'Nl-eNeni

cd OO ^t(N(NOQOOOOOONr^QCNQO(N(NONONr-Hen ~8i
OOOOOOOOOOoooOOOOOOooo^encN
O OO OO OO OO OO OO OO OO OO O (NO s-s :
w O? 3
t3 C X5
*3^
c o?
HOcNONCNO^HHunin--Hr^ONO'--l(N--HVO'-Hr--fr:* r < -S g -^
^ON'HOQOOOOO^oOHmO?KNcn^iriNowO
OOOOOOOOOO-hOOOOOOOOh^^O
O ????????????????????
wwww!www1w!ww end
^ ?3 J3
,w 8 .2
'S (N c?
??Cdu'&
2 *" ?h
fe <
O
MO. ^ 00 W)
cd "cd
> fe u i?
W) Sc c?^u
4) es bh ??h J' SB
(U p
?U 8u
1 ^ k, < QSO?

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. /., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 114 A. Chakrabarti, K. Singh and I. Mahmood

developed institutional environments and harm economy-wide shock, particularly in less devel
performance in more developed institutional envi oped institutional environments. Table 6 shows
ronments. Firms affiliated with business groups do that diversification was not related to performance
not gain from diversifying, except in less devel for the full sample. Diversification only had a neg
oped institutional environments, with the exception ative impact on change in performance during an
of a positive diversification-performance associa economy-wide shock in Japan.5 Diversification did
tion for group-affiliated firms in Singapore. This not affect performance in the other countries dur
highlights anomalous results for Singapore in sev ing an economy-wide shock. We tested the sen
eral areas, which we discuss below. Our general sitivity of results in Table 5 to the inclusion of
interpretation is that diversification only improves firms' foreign sales, but found results substantially
performance in substantially underdeveloped insti unchanged.
tutional environments. Outside of these environ
ments, diversification generally leads to poorer
Group affiliation
performance. These results support Hypothesis 1.
We repeated the analysis of Table 4 using data
Table 7 replicates Table 6, evaluating the impact of
for the shock period (1997 and 1998), but found
group affiliation during an economy-wide shock.
that results differed substantially from those re
Results at this level of analysis show the diversi
ported above. This supports the need for specific
fication-performance association differed between
investigation of the diversification-performance
group-affiliated and independent firms, and across
relationship during economy-wide shocks, the sub
ject of our next analyses.
5 The three industry and country control variables?change in
industry market capitalization, change in industry ROA, and
Hypotheses 2 and 3: Economy-wide shocks change in GNI per capita?were highly correlated for the 2-year
window of this analysis. We therefore dropped the year dummy
and institutional contexts
and change in industry market capitalization for the models in
Table 6. However, we found our results to be generally stable
Hypotheses 2 and 3 predict that diversification is when we substituted the dropped variables for change in industry
associated with performance declines during an ROA and change in GNI per capita in Tables 6 and 7.

Table 6. Effects of diversification on ROA during an economy-wide shock

Full sample Singapore Japan Korea Malaysia Thailand Indonesia

Entropy -0.002 0.029 -0.012*** 0.012 -0.004 -0.004 -0.032
(0.006) (0.020) (0.005) (0.012) (0.019) (0.043) (0.114)
Firm size (log) -0.002 -0.010 0.007*** -0.012*** 0.006 -0.012 0.009
(0.002) (0.008) (0.001) (0.004) (0.004) (0.009) (0.017)
Debt ratio -0.002*** -0.011 -0.001*** -0.003 -0.008 -0.001 -0.003
(0.000) (0.011) (0.000) (0.002) (0.009) (0.001) (0.002)
Current ratio 0.002* -0.003* 0.003* -0.006 0.008** 0.002*" 0.009*"
(0.001) (0.002) (0.002) (0.005) (0.003) (0.001) (0.003)
Firm age -0.001*** -0.001 -0.001*** -0.001** -0.003*" 0.000 0.000
(0.000) (0.001) (0.000) (0.001) (0.001) (0.001) (0.002)
Change in industry ROA 0.428** 0.058 -0.027 -0.181 0.147 -0.146 -0.451
(0.169) (0.262) (0.210) (0.486) (0.277) (0.782) (1.599)
Change in GNI per capita -3.162*** 0.205 0.031 -15.170*** -6.205 -17.446 -156.230
(0.925) (3.080) (0.788) (4.457) (13.783) (34.460) (289.018)
Constant 0.154*** 0.168 -0.048 0.422*** -0.105 0.251 -0.023
(0.038) (0.111) (0.036) (0.072) (0.104) (0.157) (0.325)
N 5,334 364 2,151 1,658 647 276 242
F 9.95*** 2.57*** 15.40*** 23.16*** 3.19*** 2.86*** 5.90***
Adjusted R2 0.030 0.022 0.066 0.036 0.112 0.048 0.002

Dependent variable is log ROA.

Data are for 1997 and 1998.
Indicator variables for five countries (for full sample model) and 18 industries for all models, not shown because of space constraints.
OLS regression with robust standard errors in parentheses; *** p < 0.01; ** p < 0.05; * p < 0.1

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 115

countries. Group affiliation accentuated perfor country's relatively developed institutional envi
mance declines during an economy-wide shock ronment.6 Independent firms in Singapore suffered
for group-affiliated firms in Japan and Malaysia. poorer performance if they diversified, consistent
Once again, Singapore was an exception, as with Hypothesis 1, and did not improve or suf
diversification improved group-affiliated firms' fer poorer performance during the economy-wide
performance during the shock. There were no shock. In contrast, and contrary to hypotheses,
diversification-related performance benefits in any group-affiliated firms' diversification was associ
other group-affiliated or independent firms in any ated with improved performance both outside and
country. during the economy-wide shock, a result not repli
In general, results in Tables 6 and 7 only cated in any other context. One possible expla
support Hypothesis 2 for Japan. Hypothesis 3, nation is provided by Khanna and Rivkin (2001),
which predicted that diversification would lead who found that group affiliation improved prof
to greater performance declines in less developed itability in relatively developed emerging econ
institutional environments, was not supported. omy capital markets. Singapore's capital mar
These results show that diversification does not kets were rated as among the most developed
moderate firm performance during economy-wide globally during the period of our study (World
shocks. Bank, 1998). A complementary explanation fol
lows from heavy government participation and
Evaluating results of Tables 5 and 7 collec
tively allows comparison of the diversification extensive MNC presence in the Singapore econ
omy (von Alten, 1995). Firms affiliated with
performance association across a period of rela
government-linked business groups comprised 37
tive stability and economy-wide shock. Our com
percent of our group-affiliated sample, while a fur
parison focuses on the general direction of the
ther 17 percent were subsidiaries of large, estab
diversification-performance relationship, limiting
lished foreign MNCs, some of which were busi
the risks of comparisons across different mod
ness groups. Firms affiliated with government busi
els. With the exception of Singapore, firms that
ness groups or MNCs may have enjoyed sub
gained from diversifying in stable periods did not
stantial reputation and resource access advantages
enjoy improved performance during the economy
in Singapore's small and rapidly growing econ
wide shock, but were able to avoid performance omy, with improved outcomes for diversification
declines. Similarly, with the exception of group in non-shock periods. Improved performance dur
affiliated firms in Japan, no firms suffered nega ing the economy-wide shock may have resulted
tive diversification-performance associations both from a 'flight to quality' trend, which is supported
outside and during shocks. Non-group firms in Sin by Singh and Ang (1999), who demonstrate that
gapore and top 30 group-affiliated firms in Korea government businesses in Singapore are unusu
had a negative diversification-performance asso ally well managed and successful. Government
ciation during the non-shock period but did not linked firms in our other sample countries do not
suffer from such diversification during the shock. enjoy similar reputations. In addition, the sub
This suggests that diversification's impact on per stantial government efforts to prime the economy
formance does not in general endure across a rad during the economy-wide shock may have chan
ically altered economic environment. neled greater resources through its affiliated firms.
Control variables generally had the expected Finally, Singapore's small and open economy is
impact across most analyses, though with varia attractive to MNCs primarily as a manufacturing
tion across countries and group affiliation. With base for exports, which may have moderated the
one exception, all models had significant though effects of the shock for MNC-affiliated firms.
varying levels of explanatory power. In general, results show that diversification only
improved firm performance in less developed insti
tutional environments. The ideal outcome?for
diversification to improve performance during the
The Singapore anomaly non-shock period and during the economy-wide

The results for Singapore's group-affiliated firms 6 We thank an anonymous reviewer for highlighting the anoma
represent an interesting anomaly, given the lous nature of the results for Singapore.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 116 A. Chakrabarti, K. Singh and I. Mahmood

GP menoocNoocnONin* ^O
O o _ooNr^r-r--ONiorMO^
Z bb dd?dddddo?NNH\?od r^d
, w , w , w w w , w^? w
Iw
o cm rt- en (NoN^(N^^mr-vOr-int^* On
O"3-O00 O
CN CN
O OOOOO^ten-d-oooocMUcom
O ?O??OOOOOO?^<N(NOO ONO
! w w i w w w wy? 00 | ?
'-HONOOiO'-'^eniO'-H^votnr^enm^o* t^
0^r-Hr-HOOOOOOr^^tunooON(N^oooo
OOOOOOOOOOenOwnCNCNCMcNiOO
<h bu ddddddddddd?^(NcNod end
w ! w ! w w w ! w^ ^ w
cNvoooen^OencNr-^
OOOOOOOO OO'sOONiOOO^O^gcNOO
podododo?o?oen^f-'?o
^ encNirir^r^cNenen^^t^t^cNr^ONr^t en
^CNOOOO'-?OOOen-H00CNV0'-HT-HOT^
OOOOOOOOOOrtenooooO^un^^
dodododododor-'iodo tj-'o

c r^o^or^ON^Tfr^^^'sf;ooo^oin^T4.* ^
G<D
? OOOOOOOOOOunTt^OO-HenUoO
?????????? ? d ^' ?ri d d ' "" I

Oh
I

<D
T3
oo ^.-Hoooooooo^HONt^ovD^oSr-m
?????????????cn??^^?
OOOOOOOOOOCNCNONTttNORoO
#C
Z bb
G
cd w I w I w I w I w w7 w w ^
OOOONencNooTj-eN'-Hr^ON^ooocMovenio
(NMOOOOO^OOO\fSHOiO(N^oo>n
O fe OOOOOOOOOO^H^vO^^HCNVj-t^O
dododododo-H*?in(Ndo -? ?
2 OONenooenrtONTj-v
-HCNOOencNO^? ?enOm^oo-d-cNo '^t
D, P OOOOOOOO? lO^iOioONT-HOr^OCN
iOioooencNi^Den!-HCN'-H
O OOOOOOOOOOOOwn<NOO
O
enoenen^-OencN^O'-^CNiiooo^Or-H^? en
OOOOOOOOOOOmcNCM^O^mO
?????????????~??"~**f?
inooeNen^eNenen^o-'d-^O^OONCNvOoo* (N
-HOOOOOOOOOoomcNvo-'d-t^rnOoo
t: S
????????????^^?? ONO
OOOOOOOOOOCNenO^f-HO^I^-H

?D
vOVDVOCNCNCN^in CN^DCNmoOVDvn? vO
---OQOQOOO
OOOOOOOOw^.H^H^qoioOO
-v.^^^w ooo^osOTfoooor
<
"O
?????????????^??
! W W|W|W|W|W|W|W oo d
O
Pi
co enio^0O"d-'-Hr^cNO'-<O^en0Nio^rsi* O
<h b? OOOOOOOOOOOCNt^ ? ^^r^OO
o o d o d o d o d o d o d ?o o ?o d

P
Oh
O
r- ?< r- < _ ooenioyocN^enin'd-unr^vo_1? on
> O OOOOOOOOOOcnm^j-mcn?<<*>ooO
dodododododo^'endo vod <
O
'-3

<
O

S Up e * zo ^ k, <
Copyright ? 2007 John Wiley & Sons, Ltd.

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 117

shock?did not exist, except for group-affiliated and institutional environments, and that this vari
firms in Singapore. Diversification only improved ation substantially affects the outcomes of their
performance for firms in the least developed insti affiliated firms' diversification.
tutional environments during stable periods but We draw four overall implications. First, there
not during an economy-wide shock. Negative out are limited and specific conditions under which
comes for Korean affiliated firms in stable periods diversification is performance enhancing, espe
were associated with unchanged performance dur cially if sudden and major changes occur in the
ing the shock. And in the worst outcomes, diver environment. Second, the outcomes of diversifi
sification led to performance declines during both cation vary substantially across environments that
stable and shock periods for group-affiliated firms vary in institutional development and economic
in Japan. stability. Third, affiliation with a business group
These results are consistent with findings of a does not generally improve diversification's out
diversification discount in more developed institu comes. Fourth, and following from the above
tional environments (Montgomery and Wernerfelt, conclusions, the proposition that diversification is
1988; Ramanujam and Varadarajan, 1989), and more likely to improve performance in less devel
with the view that diversification can substitute for oped institutional environments is subject to sev
inefficient markets in less developed institutional eral contingencies.
environments (e.g., Khanna and Palepu, 1997). Debate on the outcomes of diversification is
They are also consistent with studies that suggest unlikely to be resolved without continued atten
limited benefits to diversification (Lubatkin and tion to the contingent nature of the diversifi
Chatterjee, 1994; Montgomery, 1994; Ramanujam cation-performance relationship. We have high
and Varadarajan, 1989), particularly during sys lighted three important contingencies: the insti
temic shocks. tutional environment, the stability of the broad
economic environment, and business group affil
iation. This perspective is consistent with Kogut
et al. (2002), who propose that nation-specific
DISCUSSION AND CONCLUSIONS
conditions may prevent convergence on specific
diversification outcomes, and with the view that
We find that the outcomes of diversification are
institutional and country heterogeneity will persist
influenced by institutional environments, economic (Ahmadjian and Robbins, 2005; Guillen, 2001).
stability, and business group affiliation. The out Hence, closer attention to contextual, institutional,
comes of diversification vary across institutional and country characteristics is required (Aoki, 2001;
contexts but are positive only in the most under Biggart and Hamilton, 1992; Peng, 2002), partic
developed institutional environments, and nega ularly as studies continue to find substantial and
tive in more developed environments. Diversi sustained differences in firm behavior and perfor
fication does not generally alleviate the impact mance across countries (Khanna and Rivkin, 2001;
of an economy-wide shock on firm performance. Kogut et al, 2002; Lins and Servaes, 1999).
This result provides empirical support for an Another dimension of diversification's com
unstated assumption of the diversification litera plexity rests on the likelihood that firms most
ture, that diversification does not provide substan likely to gain from diversifying?those operat
tial 'spreading of risks' benefits for firms facing ing in weak institutional environments?may lack
systemic or economy-wide shifts in economic con the resources to undertake and manage the com
ditions. More generally, our results indicate that plexities associated with such diversification. The
the outcomes of diversification are contingent on market and institutional limitations that support
stability in the economic environment. We also find diversification, such as shortage of managerial
that group affiliation often affects the outcomes talent, and financial and information inefficien
of diversification. In most cases, the outcomes of cies, may constrain firms from effectively man
diversification differ significantly, though not in a aging that diversification, particularly if external
consistent direction, between group-affiliated and conditions change rapidly. Thus, business group
non-group firms within and across countries, in affiliated firms who benefit from internal resource
both stable and shock periods. This suggests that transfers during stable periods suffer reverses when
the nature of business groups varies across country economic conditions change drastically.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. 7., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 118 A. Chakrabarti, K. Singh and I. Mahmood

Relatedly, studies of business groups and diver environments. The absence of such measures rep
sification in less developed institutional envi resents one limitation of our study.
ronments focus on environment-derived advan Economic shocks are valuable contexts for
tages, but have not adequately examined how research, serving as natural experiments for testing
these same environmental conditions may result the boundary conditions of various associations.
in high organization costs (e.g., Hannan et al., As economic shocks are common within the
2003; Williamson, 1985). Attention to organiza global economic and financial architecture,
tional cost and complexity challenges in devel examining diversification-performance and other
oping institutional contexts will improve under relationships in environments with recurring
standing of group and diversification outcomes. economic shocks may provide useful insights.
Hence, business group affiliation is not unambigu Economy-wide shocks transmit their effects, in
ously beneficial for firms, indicating conditionality large part, through impact on individual firms.
for the view that groups are relatively effective at Understanding the micro-economic impact of these
exploiting institutional weaknesses. Closer atten shocks, especially across different institutional
tion to how groups' embeddedness influences affil environments, will improve our understanding of
iated firm's ownership structures, management and the effects of macro-economic change, and of how
cost structures, diversification strategies and per firms react to such change.
formance, particularly in less studied and more
idiosyncratic institutional environments, will pro
vide useful elaboration and clarification of their ACKNOWLEDGEMENTS
roles.
Diversification remains a challenging strategy,
We thank Andrew Delios, Sea Jin Chang, Anand
even in relatively munificent circumstances of high
Swaminathan, and two anonymous reviewers for
growth in relatively weak institutional environ
valuable inputs.
ments, contexts that market failure perspectives
propose as favorable for profitable diversifica
tion. It is not clear that extensive research has
REFERENCES
improved managerial understanding of the firm
specific, complex, and conditional issues that influ Ahmadjian CL, Robbins GE. 2005. A clash of capi
ence business scope decisions. It would be use talisms: foreign shareholders and corporate restructur
ful to specify and communicate to managers the ing in 1990s Japan. Academy of Management Meet
ex ante conditions for successful diversification. ings, Honolulu.
Alston LJ, Eggerston T, North DC (eds). 1996. Empirical
From a policy perspective, our results support
Studies in Institutional Change. Cambridge University
recent questioning (e.g., Shin and Chang, 2003) Press: Cambridge, U.K.
of the effectiveness of development strategies cen Aoki M. 2001. Towards a Comparative Institutional
tered on resource prioritization for highly diversi Analysis. MIT Press: Cambridge, MA.
fied firms. Aron J. 2000. Growth and institutions: a review of the
evidence. World Bank Research Observer 15: 99-135.
Despite persuasive arguments for the impor
Ashby WR. 1969. Self regulation and requisite variety. In
tance of institutions (Khanna and Palepu, 1997;
Systems Thinking, Emery FE (ed). Penguin: London.
Williamson, 1985), relatively little empirical strat Backman M. 1999. Asian Eclipse: Exposing the Dark Side
egy research has adopted a comparative institu of Business in Asia. Wiley: Singapore.
tional perspective or undertaken specific investi Bergh DD, Fairbank JF. 2002. Measuring and testing
gation of their impact. One means for addressing change in strategic management research. Strategic
Management Journal 23(4): 359-366.
this gap is to conduct more research in differ
Biggart NW, Hamilton GG. 1992. On the limits of a firm
ent countries and institutional environments, not
based theory to explain business networks: the Western
necessarily as a means for 'internationalizing strat bias of neoclassical economics. In Networks and
egy,' but as a means for introducing contextual Organizations, Nohria N, Eccles RG (eds). Harvard
Business School Press: Boston, MA; 471-490.
variation. Research of this nature has the potential
Chang SJ, Hong J. 2002. How much does the business
to contribute to strategy theory, not just to improve
group matter in Korea? Strategic Management Journal
understanding of the contexts in which such theory 23(3): 265-274.
is applied. A prerequisite is theoretically derived Corsetti G, Pesenti P, Roubini N. 1999. The Asian cri
and empirically validated measures of institutional sis: empirical evidence and policy debate. In The

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
 Diversification and Performance 119

Asian Financial Crisis: Causes, Contagion and Con Khanna T, Palepu K. 1997. Why focused strategies may
sequences, Agenor PR, Miller M, Vines D, Weber A be wrong for emerging markets. Harvard Business
(eds). Cambridge University Press: Cambridge, UK; Review 75(4): 41-51.
127-162. Khanna T, Palepu K. 1999. Policy shocks, market
Corsetti G, Pesenti P, Roubini N. 2001. Fundamental intermediaries, and corporate strategy: evolution of
determinants of the Asian crisis: the role of financial business groups in Chile and India. Journal of
fragility and external imbalances. In Regional and Economics and Management Strategy 8: 271-310.
Global Capital Flows: Macroeconomic Causes and Khanna T, Rivkin J. 2001. Estimating the performance
Consequences, Ito T, Krueger AO (eds). University of effects of business groups in emerging economies.
Chicago Press: Chicago, IL; 11-41. Strategic Management Journal 22(1): 45-74.
DessGD, Gupta A, Hennart J, Hill CWL. 1995. Con Kock C, Guillen MF. 2001. Strategy and structure
in developing countries: business groups as an
ducting and integrating strategy research at the interna
evolutionary response to opportunities for unrelated
tional, corporate and business levels: issues and direc
diversification. Industrial and Corporate Change
tions. Journal of Management 21: 357-393.
10(1): 77-113.
Donaldson L. 2000. Organizational portfolio theory: per
Kogut B, Walker G, Anand J. 2002. Agency and
formance driven organizational change. Contemporary
institutions: national divergences in diversification
Economic Policy 18: 386-396.
behavior. Organization Science 13(2): 162-178.
Dornbusch R, Park YC, Ciaessens S. 2000. Contagion:
Lincoln JR, Gerlach MH, Ahmadjian CL. 1996. Keiretsu
understanding how it spreads. World Bank Research networks and corporate performance in Japan.
Observer 15: 177-197.
American Sociological Review 61: 67-88.
Ghemawat P, Khanna T. 1998. The nature of diversified Lins K, Servaes H. 1999. International evidence on the
business groups: a research design and two case value of corporate diversification. Journal of Finance
studies. Journal of Industrial Economics 46(1): 54: 2215-2239.
35-62. Lubatkin M, Chatterjee S. 1994. Extending modern port
Grant RM, Jammine AP, Thomas H. 1988. Diversity, folio theory into the domain of corporate diversifica
diversification, and profitability among British man tion: does it apply? Academy of Management Journal
ufacturing companies. Academy of Management Jour 37(1): 109-136.
nal 31: 771-801. Markides CC. 1992. Consequence of corporate refocus
Greenwood R, Hinnings CR. 1996. Understanding radi ing: ex-ante evidence. Academy of Management Jour
cal organizational change: bringing together the old nal 35: 398-412.
and the new institutionalism. Academy of Management Markides CC, Williamson PJ. 1996. Corporate diversifi
Review 21: 1022-1054. cation and organizational structures: a resource based
Guillen MF. 2000. Business groups in emerging view. Academy of Management Journal 39: 340-367.
economies: a resource-based view. Academy of Matsusaka J. 1993. Takeover motives during the
Management Journal 43(3): 362-380. conglomerate takeover wave. RAND Journal of
Economics 24: 357-379.
Guillen M. 2001. The Limits of Convergence: Globaliza
tion and Organizational Change in Argentina, South Montgomery CA. 1994. Corporate diversification. Jour
Korea, and Spain. Princeton University Press: Prince nal of Economic Perspectives 8: 163-178.
ton, NJ. Montgomery CA, Wernerfelt B. 1988. Diversification,
Hamilton GG, Biggart NW. 1988. Market, culture and Ricardian rents, and Tobin's q. RAND Journal of
Economics 19: 623-632.
authority: a comparative analysis of management and
Newman K. 2000. Organizational transformation during
organization in the Far East. American Journal of
institutional upheaval. Academy of Management
Sociology 94: 252-294. Review 25: 602-619.
Hannan MT, Freeman JH. 1984. Structural inertia and
Ocasio W. 1997. Towards an attention based view of the
organization change. American Sociological Review
49: 149-164. firm. Strategic Management Journal, Summer Special
Issue 18: 187-206.
Hannan MT, Polos T, Carroll G. 2003. Cascading
Palepu K. 1985. Diversification strategy, profit perfor
organizational change. Organization Science 14(5): mance and the entropy measure. Strategic Manage
463-482.
ment Journal 6(3): 239-255.
Hausman JA. 1978. Specification tests in econometrics. Palich LE, Cardinal LB, Miller CC. 2000. Curvilinearity
Econometrica 46: 1251-1272.
in the diversification-performance linkage: an exam
Hoskisson RE, Hitt MA. 1994. Downscoping: How to ination of over three decades of research. Strategic
Tame the Diversified Firm. Oxford University Press: Management Journal 21(2): 155-174.
New York.
Peng MW. 2002. Towards an institution-based view of
Hoskisson RE, Eden L, Lau CM, Wright M. 2000. business strategy. Asia Pacific Journal of Management
Strategy in emerging economies. Academy of 19: 251-267.
Management Journal 43(3): 249-267. Raghunathan SP. 1995. A refinement of the entropy
Jacquemin DB, Berry CH. 1979. Entropy measure of measure of diversification: toward definitional and
diversification and corporate growth. Journal of computational accuracy. Journal of Management 21:
Industrial Economics 27: 359-369. 989-1002.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms
120 A. Chakrabarti, K. Singh and I. Mahmood

Rajagopalan N, Spreitzer GM. 1997. Toward a theory Thomsen S, Pedersen T. 2000. Ownership structure
of strategic change: a multi-lens perspective and and economic performance in the largest European
integrative framework. Academy of Management companies. Strategic Management Journal 21(6):
Review 22: 48-79. 689-705.
Ramanujam V, Varadarajan P. 1989. Research on corpo von Alten F. 1995. The Role of Government in the
rate diversification: a synthesis. Strategic Management Singapore Economy. Peter Lang: Frankfurt.
Journal 10(6): 523-551. Wan C, Hoskisson RE. 2003. Home country environ
Shin J-S, Chang H-J. 2003. Restructuring Korea Inc. ments, corporate diversification strategies, and firm
Routledge-Curzon: London. performance. Academy of Management Journal 46:
Siggelkow N, Levinthal DA. 2003. Temporarily divide to 27-45.
conquer: centralized, decentralized, and reintegrated White S. 2002. Rigor and relevance in Asian management
organizational approaches to exploration and adapta research: where are we and where can we go? Asia
tion. Organization Science 14(6): 650-669. Pacific Journal of Management 19: 287-352.
Singh K, Ang SH. 1999. Governments in business: an Williamson OE. 1985. The Economic Institutions of
empirical analysis of the strategies and success Capitalism. Free Press: New York.
of government linked corporations in Singapore. Wooldridge JM. 2002. Econometric Analysis of Cross
Best Paper Proceedings, Academy of Management: Section and Panel Data. MIT Press: Cambridge, MA.
Chicago, IL; C1-C6. World Bank. 1998. World Development Indictors. World
Singh K, Yip G. 2000. Strategic lessons from the Asian Bank: Washington, DC.
crisis. Long Range Planning 33: 706-729.

Copyright ? 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., 28: 101-120 (2007)
DOI: 10.1002/smj

This content downloaded from 192.76.8.93 on Wed, 22 Nov 2023 06:14:41 +00:00
All use subject to https://about.jstor.org/terms