The Persistence and Implications of Japanese Keiretsu Organization

Author(s): J. McGuire and S. Dow

Source: Journal of International Business Studies, Vol. 34, No. 4 (Jul., 2003), pp. 374-388
Published by: Palgrave Macmillan Journals
Stable URL: http://www.jstor.org/stable/3557181 .
Accessed: 23/03/2011 07:13

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=pal. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

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 support@jstor.org.

Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve and extend access to Journal of
International Business Studies.

http://www.jstor.org
 X%.'^1-/~~~ of International
lJournal Studies
Business (2003)34,374-388
C-2003 PalgraveMacmillan
Ltd. Allrightsreserved0047-2506 $25.00
www.jibs.net

The persistence and implicationsof Japanese

keiretsu organization

J McGuirel and S Dow2 Abstract

This paper examines two important research questions: (1) To what extent
1JohnMolsonSchoolof Business,Concordia have recent economic and regulatory changes influenced the Japanese inter-
Montreal,Canada;2Universite
University, du corporate network? (2) Have the patterns of firm performance fostered by the
Quebeca Montreal,Ecoledes sciencesde la Japanese inter-corporate system remained stable, or has there been an
gestion,Montreal,Canada evolution to more 'North American' performance profiles? Using hierarchical
regressions to compare patterns of networking and firm performance with data
Correspondence: from the periods 1987 to 1991 and 1992 to 1997, we find little evidence to
Dr J McGuire,Professor of Management,
Concordia University,Montreal, Quebec, support the existence of significant changes in the patterns and implications of
Canada. Japanese industrial organization. We conclude that the keiretsu system
Tel: +1 (514) 848 2917; remained strongly in place throughout the first half of the 1990s. Nevertheless,
Fax: + 1 (514) 848 4292; we suggest that the continued move toward globalization of capital markets in
E-mail:jeanm@jmsb.concordia.ca Japan and ongoing regulatory change may potentially impact networking and
performance implications in the 21st century.
journal of International Business Studies (2003) 34, 374-388. doi:10.1057/palgrave.
jibs.8400038

Keywords: keiretsu;industrialgroupings

Introduction
Although the Japanese keiretsu system has long been considered a
source of strategic advantage for Japanese firms (Sheard, 1991,
1994a; Williamson, 1991a,b), scholars have recently begun to
question the continued viability of this system. First, the stable
shareholdings and close banking ties that represent the corner-
stones of keiretsu links may themselves be threatened. Globaliza-
tion of financial markets, deregulation of Japanese securities
markets and the Asian economic downturn may have generated a
more 'North American' investment context that emphasizes
market performance (Useem, 1998; Yoshikawa and Phan, 2001a).
These changes could suggest growing 'convergence' between the
US and Japanese models of industrial organization (Roe et al., 1993)
in which the traditional stakeholdermodelof the Japanese firm may
be evolving toward more North American shareholder models
(Thomas and Waring, 1999). In essence, firms may increasingly
respond to the demands of global investors rather than their
traditional local stakeholders.
Received: 14 August 2000
Second, insulation from financial and market pressures provided
Revised: 19 February2003 by bank ties and stable shareholdings may hold fewer advantages
Accepted: 23 February 2003 in the context of economic change and globalized financial and
Online publication date: 22 May 2003 product markets (Johnston and McAlevey, 1998). Indeed,
 lapanese
I I~ keiretsu
~---- oraanization
Ito (1997) and Dow and McGuire (1999) suggest
that the characteristics of the Japanese industrial
system that have contributed to the growth of the
Japanese economy may now act as constraints.
This paper examines changes in Japaneseindustrial
organization between the period 1987 to 1991 and
1992 to 1997. It addresses two important research
questions: (1) To what extent have recent economic
and regulatory changes influenced the Japanese
inter-corporate network? (2) Have the patterns of
firm performance fostered by the Japanese context
remained stable, or has there been an evolution to
more 'North American'performance profiles?
These issues have considerable significance for
the field of international business. There has been
increasing recognition that industrial organization
and corporate governance have a significant impact
on the behavior and performance of firms (Suzuki
and Wright, 1985; Frankel, 1991; Roe et al., 1993;
Ahmadjian and Robbins, 1999; Thomas and
Waring, 1999; Ahmadjian and Lincoln, 2001).
Useem (1998) argues that globalization of equity
markets has created pressures to open financial
markets to a wider range of financial stakeholders.
In essence, international business pressures may
provide a key channel in the evolution of national
regulatory contexts. Further, aspects of Japanese
industrial organization are found in many other
contexts. For example, strong banking ties are
found in a number of economies (such as
Germany), as are inter-corporate groupings
(Steers et al., 1989; Guillen, 2000; Khanna and
Palepu, 2000; Khanna and Rivkin, 2001). An
understanding of the strategic implications of
these inter-corporate linkages is therefore an
important aspect of the context of international
business.
Theoretical framework: Japanese inter-
corporate networks
Japanese industrial organization differs from that
found in North America in that it is characterized
by a complex web of inter-corporate networks. This
study focuses on two aspects of the Japanese
context. First, the structure of firm ownership and
banking relations found in Japan differs signifi-
cantly from that typically found in North America.
Second, keiretsu groupings - well-established net-
works of horizontally linked firms - play an
important role in the Japanese economy. The
following sections will outline the basic character-
istics of these ties and their implications for firm
strategy.
I McGuire and S Dow *^
375
Equity ties
The ownership of Japanese firms is dominated by
institutional and corporate holdings. In 1990,
financial institutions and corporate investors held
approximately 71% of outstanding shares in
Japanese firms. This compares with approximately
51% individual ownership, with higher institu-
tional investment in the US (Sheard, 1994d;
Prowse, 1995). Moreover, the role of institutional
holdings differs in the two contexts. Whereas in the
US certain institutional investors such as pension
funds play an active role in corporate governance,
their role is more limited in Japan owing to
regulations restricting the activities of non-bank
institutional investors. Perhaps as a result of these
restrictions, institutions hold only approximately
16% of household assets in Japan as compared with
approximately 33% of household assets in the US
(Gibson, 1998). As will be discussed in more detail
below, however, bank holdings represent a signifi-
cant ownership block in Japan.
The extent and type of input available to share-
holders also differs between Japan and North
America. North American owners typically exert
formal, arm's-length input into corporate govern-
ance through the board of directors, shareholder
initiatives, and the like. In particular, institutional
investors act as external monitors, and play an
active role in corporate governance.
In Japan, however, inter-corporate and
bank shareholders act as quasi-insiders, linked
to the focal firm through other types of ties
(buyer-seller relationships, bank ties, board
interlocks, etc.). Gibson (1998) estimates that 38%
of the shares in large Japanese firms can be
classified as 'quasi-inside' holdings, including
the bank holdings discussed below. Also in
contrast to the US, non-bank institutional investors
play a relatively passive role in formal corporate
governance. In the context of this lack of formal
corporate governance activity, Japanese share-
holders rely on the less formal input made possible
by their multiple ties with other shareholders
and the firm. Thus these quasi-inside shareholders
may have significant input into firm decision-
making.
Further, these quasi-inside stable shareholders
provide Japanese firms with capital market stability,
in which shares are seldom traded, particularly to
outside owners. For example, over 82% of the 10
largest shareholdings in Gerlach'sJapanese sample
remained stable over the period 1980-1984, as
compared with only 23% in his US sample
Journal of International Business Studies
\I/
* . TX' Japanese keiretsu organization
376
(Gerlach, 1992, 76). This figure remained relatively
stable over the longer period 1969-1986.
Holdings are also more likely to be reciprocal in
Japan. These balanced and long-term interests
contrast with the control implied by one-direc-
tional ownership interests that are prevalent in
North America. In essence, the motives of such
stable reciprocal shareholders may focus on main-
taining business or strategic relationships and
providing mutual support, rather than seeking
capital gain. As a result, the emphasis on share-
holder well-being and the prioritization of financial
claimants found in North America might not be
applicable in Japan (Sheard, 1994d; Phan and
Yoshikawa, 2000). Thus the North American share-
holder wealth maximization model may not be
applicable to the Japanese context (Yoshikawaand
Phan, 2001a).
Bank ties
Bank ties are another aspect of Japanese industrial
organization. Japanese firms rely heavily on bank
debt for financing needs. They are very often linked
to a small number of banks, led by a 'main bank' for
a significant portion of their financing needs (Roe
et al., 1993; Corbett, 1994; Hoshi, 1994). These
banking relationships, particularlywhen combined
with equity holdings and the other ties described
above, encourage banks to provide stable financing.
This provides more ready access to funding, and
reduces the need to carryreserves of long-term debt
or liquidity (Hoshi, 1994; Weinstein and Yafeh,
1998) compared with the situation for non-main
bank firms. To illustrate, bank lending group
members are less dependent upon traditional
financial criteria (Weinstein and Yafeh, 1998). This
has three important implications (Gilson and Roe,
1993). First, it provides greater information and
transparency to facilitate monitoring of the firm.
Second, equity positions reinforce lending ties and
provide incentives for lenders to work with the firm
to resolve any financial difficulties (Sheard, 1994c).
Finally, dual positions provide additional avenues
for input into firm decisions and additional means
of intervention if necessary. Thus Yoshikawa and
Phan (2001b) note that the claims of financial
stakeholders are less strictly prioritized in Japan.
Owing to both legal regulation and custom, US
banks and financial institutions maintain an arm's-
length relationship with firms, and are active in
corporate governance. They generally do not hold
both debt and equity. Without the history of strong
ties with financial intermediaries found in Japan,
journal of International Business Studies
I McGuire and S Dow
US firms enjoy greater strategic and financial
autonomy from major institutional actors. This
independence is reinforced by the comparatively
smaller role played by banks and other financial
intermediaries in the US context as compared with
Japan (Roe et al., 1993). Theoretically, Weinstein
and Yafeh (1998) argue that this financial indepen-
dence allows US firms to lower their cost of capita
as compared with Japanese main bank firms.
Nevertheless, this system is not without its dis-
advantages. US firms do not benefit from the patient
capital associated with Japanese banking relation-
ships. Further, banks and financial institutions
cannot provide the in-depth monitoring made
possible by multi-layered ties to the firm. In
essence, the North American system represents a
trade-off between arm's-length regulatory and
market monitoring and network monitoring
(Yoshikawa and Phan, 2001b). Although recent
corporate governance failures have challenged the
North American system as a 'Swiss cheese of
loopholes' (Rossant et al., 2002), corporate govern-
ance advocates have continued to call for the North
American model of more stringent regulation,
disclosure, and arm's-length transactions (Barber,
2002; Bruce, 2002; Harnischfeger, 2002; Major and
Lebert, 2002; Parker,2002; Stewart, 2002).
Keiretsu organization
In addition to the overall patterns of ownership and
financing described above, keiretsu groupings play
a critical role in the Japanese economy. There are
six commonly recognized horizontal keiretsu:
Mitsubishi, Mitsui, Sumitomo, Fuji, Sanwa, and
Dai Ichi Kango. Several keiretsu date back to before
the 19th century. Others date from the post-World
War II period. These long-standing historical rela-
tionships reinforce the in-group debt and equity
holdings of keiretsu firms. The ownership structure
of firms is dominated by group holdings, which
range from 23 to 42% (Gerlach, 1992, 119). Major
shareholders are typically the main bank and its
affiliates and other associated firms (Hoshi, 1994).
Reciprocal shareholdings are more prevalent
among keiretsu firms (Gerlach, 1992; Prowse,
1992; Sheard, 1994a). Perhaps as a result of the
role of group shareholdings, shares in keiretsu firms
are less frequently traded. When traded, they are
likely to be placed with a previous shareholder,
usually a keiretsu member (ohnston and McAlevey,
1998).
Other personnel ties complement equity ties, and
transactional links reinforce the mutual interest
 laDanese keiretsu oraanization
---I
group stakeholders (Berglof and Perotti, 1994;
Sheard, 1994d; Kaplan and Minton, 1994; Hwang
and Kim, 1998; Morck and Nakamura, 1999). As a
result, keiretsu ties do not necessarily imply large
individual stakes. Rather, keiretsu members as a
whole dominate the ownership structure. These
multiple ties allow for a dispersed yet concentrated
ownership network.
Keiretsu firms rely more heavily upon bank
financing than do non-keiretsu firms (Hoshi,
Kashyap, and Sharfstein, 1991; Gerlach, 1992;
Prowse, 1992). Typically, the main bank accounts
for 10-20% of a firm's financial borrowing (Flath,
1993, 251). Total group lending can run as high as
63% (Gerlach, 1992, 126). These lending ties are
also more stable (Gerlach and Lincoln, 1992), and
are reinforced by other ties to the firm. For
example, equity holdings are significant. In 57 of
the 133 keiretsu firms examined by Prowse (1990),
the largest shareholder was also the largest debt
holder. On average, the top five debt holders owned
18.2% of outstanding shares. Bank representatives
often sit on corporate boards, and offer assistance
in times of financial distress (Berglof and Perotti,
1994; Sheard, 1994b; Morck and Nakamura, 1999).
Our discussion thus far suggests that both
keiretsu and non-keiretsu firms differ from North
American firms in terms of their ownership char-
acteristics and banking ties. Both categories of
Japanese firm are embedded in a network of
interrelated and stable financial stakeholders. How-
ever, the stronger, more intertwined ownership
structure and closer ties with banks suggest that
differences between keiretsu firms and US firms
may be significant.
Benefits of Japanese industrial organization
This stable network of financial stakeholders has
traditionally provided Japanese firms, particularly
keiretsu firms, with important strategic advantages.
Although interrelated, it is useful to discuss these
benefits in terms of more ready access to financing,
risk reduction, and product-market benefits. Quasi-
inside reciprocal shareholdings insulate the firm
from market pressures for short-term performance
(Nakatani, 1984; Prowse, 1992; Sheard, 1994a).1
Close relationships with banks and other lenders
provide the firm with ready access to financing
(Nakatani, 1984; Frankel, 1991; Sheard, 1994b). As
a result, Japanese firms, particularly keiretsu firms,
may have greaterfinancial flexibility than do North
American firms.
I McGuire and S Dow
-. - E-
*
377
Further,risk is reduced for the firm, its creditors,
and its shareholders. Firm risk is attenuated by
stable shareholdings and the patient capital pro-
vided by both equity and credit. Stable share-
holding arrangements tend to insulate the firm
from market volatility. Banks, as both shareholders
and lenders, are likely to work with financially
troubled firms by providing managerial assistance,
extending payment terms, or purchasing inventory
(Aoki, 1994; Sheard, 1994c). The reciprocal mon-
itoring by a closely connected set of financial
stakeholders described earlier opens the firm to
wide-ranging scrutiny, which reduces risk for
creditors (Roe et al., 1993; Sheard, 1994d; Kim and
Limpaphayom, 1998; Morck and Nakamura, 1999).
Risk is further reduced by the assistance given to
troubled firms by other group members (Hoshi,
1994; Sheard, 1994c).
The product-market benefits of keiretsu organiza-
tion stem from more ready access to suppliers, and
from access to a stable domestic market (Gerlach,
1992; Hundley and Jacobson, 1998). Although
keiretsu firms do not limit their purchases to
keiretsu partners, in-group transactions are signifi-
cant (Gerlach, 1992).
Strategic and performance implications of
Japanese inter-corporate networks
This insulation from short-term financial pressures,
and the dominance of a stable debt/equity network,
have significant implications for the strategic and
performance profiles of Japanese firms. Simply
stated, the performance and strategic objectives of
Japanese firms may reflect their stakeholder (in
contrast to shareholder) orientation (Yoshikawa
and Phan, 2001b; Barber, 2002; Bruce, 2002;
Harnischfeger, 2002; Major and Lebert, 2002;
Parker,2002; Phan and Yoshikawa, 2000; Stewart,
2002). As shareholders in Japanese firms benefit not
only from their equity holdings, but also from
their other ties to the firm, Japanese firms are
frequently characterized as favoring long-term
development over short-term financial perfor-
mance (Nakatani, 1984; Aoki, 1990; Williamson,
1991a; Gibson, 1998). Strong debt-equity ties with
banks may also promote more conservative strate-
gies that limit firm performance and growth
(ohnston, 1995; Kim and Hoskisson, 1996; Dow
and McGuire, 1999).
Empirical studies have supported the conclusion
that, although both categories of Japanese firms
exhibit weaker financial performance and growth
than US firms, these differences are most
Journal of International Business Studies
* lapanese Ikeiretsu orqanization
?~~~~~~~~~~I-.---.
378
pronounced between keiretsu and US firms. Empiri-
cal support for lower profitability among keiretsu
firms is fairly consistent (Caves and Uekasa, 1976;
Nakatani, 1984; Gerlach, 1992; Hundley and Jacob-
son, 1998; Dow and McGuire, 1999).
This more relaxed monitoring may also imply
fewer pressures toward cost containment and
efficiency (Yoshikawa and Phan, 2001a). The
multiple ties among firms suggest that network
maintenance objectives would take precedence
over short-term cost containment for both
categories of Japanese firms as compared with US
firms (Yoshikawa and Phan, 2001a). Further,
maintenance of network ties may imply additional
costs for Japanese firms. Any such additional cost
would be particularly relevant to the more closely
intertwined keiretsu firms. Although certain of
these costs may be offset by reductions in coordina-
tion costs and increased flexibility, empirical evi-
dence supports the additional costs of keiretsu
membership. Hundley and Jacobson (1998), for
example, document lower sales per employee
among keiretsu firms as compared with indepen-
dent firms. Nakatani (1984) observes higher
levels of employee compensation among keiretsu
firms.
Finally, Japanese industrial organization may
influence growth characteristics. Japanese firms'
more conservative strategies and access to a more
stable domestic market may reduce incentives to
pursue external growth, particularly for keiretsu
firms. Serving these internal markets may take
priority over exploiting growth opportunities
(Kim and Hoskisson, 1996). Kim and Hoskisson
(1996) further suggest that keiretsu firms have more
limited autonomy, whereas independent firms
enjoy more freedom in strategic action. Nakatani
(1984) finds lower growth rates among keiretsu
firms. Despite preferential access to capital,
Weinstein and Yafeh (1998) are unable to verify
differences in the growth rates of bank-affiliated
and independent firms. These authors interpret
their findings as indicative of the influence of the
main bank toward more conservative investment
policies than would otherwise be the case.
This line of reasoning also suggests that the
foreign involvement of Japanese firms, particularly
keiretsu firms, may be less than that for North
American firms. Arguing that keiretsu ties reduce
the need to rely on external markets, Hundley and
Jacobson (1998) and Geringer et al. (2000) discern
lower export performance among keiretsu firms as
compared with independent firms.
Journalof InternationalBusinessStudies
I McGuireand S Dow
Keiretsu ties in a changing economy
Ongoing changes in the economic and institutional
infrastructure of Japan have led researchers to
question the continued viability of keiretsu organi-
zation. Gibson (1998, 1), however, observes that:
Although the Japanese corporate governance system played
a key role in Japan's outstanding economic performance of
the 1960s and 1970s, times have changed and the Japanese
corporate governance system has not.
Although much of the earlierliteraturefocuses on
the benefits of keiretsu organization, scholars have
begun to highlight the potential costs of this
system (Gerlach, 1992; Kim and Hoskisson, 1996;
Weinstein and Yafeh, 1998; Dow and McGuire,
1999).
Evolution of Japanese equity markets
Throughout the decade of the 1990s continued
revisions to Japan's Commercial Code heightened
the transparency of the Japanese capital market and
thus contributed toward its globalization. These
changes came on the heels of regulatory adjust-
ments that occurred in the 1980s that included
lifting of restrictions on institutional investment,
liberalization of regulations regarding pension
funds, increased openness to foreign investment,
and the easing of entry of foreign institutional
investors into the Japanese market (Gibson, 1998).
Specific examples of regulatory change during the
1990s include, but are not limited to, the following
examples. In 1993 the Code was amended to
facilitate shareholder access to accounting records,
with a view to improving the quality of the
shareholder oversight function; in 1994 the effec-
tive ban on corporate stock buy-backswas removed;
in 1997 the ban on pure holding companies was
repealed, and restrictions on ownership of stock by
financial institutions were also eased; and in 1999
enhanced disclosure procedures were put in place.
In general, changes in the last two decades of the
20th century opened the Japanese economy to
increased competition. Such changes in the Japa-
nese context have confirmed global trends toward
more open financial markets. Pressures from a
wider range of global investors may expose Japa-
nese firms to increased performance pressures
(Useem, 1998). This increased competition may
place a premium on the firm's ability to react
quickly to changing conditions and respond to the
performance expectations of global investors.
The extent to which traditional Japanese equity
relationships can persist in this setting is uncertain.
 Japanese keiretsu orqanization
The liberalization of equity regulations noted ear-
lier has encouraged foreign investment in Japan
(Weinstein and Yafeh, 1998; Ahmadjian and
Robbins, 1999). Such foreign investment, particu-
larly that of institutional investors, may place
increased performance pressures on Japanese firms.
In the context of regulatory change and economic
decline the traditional stakeholder model of the
Japanese firm may be threatened by the North-
American shareholder model (Thomas and Waring,
1999; Phan and Yoshikawa, 2000).2 The costs of
maintaining stable shareholding and banking rela-
tionships become particularly relevant in the con-
text of more competitive capital markets. Indeed,
several studies have provided evidence of increas-
ing shareholder orientation among Japanese firms
(Phan and Yoshikawa, 2000; Yoshikawa and Phan,
2001b; Yoshikawa and Gedajlovic, 2002).
The differential impact of these changes on
keiretsu and independent firms is uncertain.
Keiretsu membership may isolate firm manage-
ment from competitive pressures (ohnston and
McAlevey, 1998). One of the traditional strengths
of keiretsu has been their role as a buffer against
market downturns, which allows the firm to
develop long-term competitive strengths. In doing
so, however, non-competitive firms may survive
with the support of more competitive partners. The
costs of supporting less competitive firms may
become increasingly burdensome in the context
of greater global competition. Faced with greater
information asymmetries and difficulties in mon-
itoring keiretsu firms, Dewenter et al. (2001) argue
that keiretsu firms may be placed at a disadvantage
in attracting outside investors. In the short term
this may serve to preserve keiretsu ties. However,
the eventual costs of reliance on group capital in
terms of more limited access to capital may lead
keiretsu firms to seek outside investment. Empirical
evidence for the persistence of equity ties among
keiretsu firms is mixed. Several studies provide
evidence for the persistence of stable shareholding
arrangements among keiretsu firms (Gerlach, 1992;
Johnston and McAlevey, 1998). Further,Johnston
and McAlevey (1998) verify the re-establishment of
cooperative shareholding arrangements, particu-
larly during more recent periods of economic
downturn. Yoshikawa and Phan (2001a) provide
anecdotal evidence that, although keiretsu firms
may be more resistant to market-model pressures,
they have increasingly adopted more market-model
performance characteristics. McGuire and Dow
(2002) also conclude that market-model inroads
J McGuire and S Dow
379
among keiretsu firms are occurring. Phan and
Yoshikawa (2000) and Yoshikawa and Gedajlovic
(2002) find that, although exposure to outside
investor pressures is associated with market-model
performance objectives, group affiliation does not
appear to influence performance characteristics.
Changing banking relationships
The role of banks in the Japanese economy has
changed in significant ways. First, the Japanese
banking industry was particularly hard hit by
declines in the Japanese equity and real-estate
markets, as well as by declines in most Asian
economies. Not only were the banks' domestic
investments in corporate loans, real estate, and
corporate equity jeopardized, but Japanese banks
were also heavily invested in other declining Asian
markets. This economic situation may have
severely strained the ability of Japanese financial
institutions to insulate Japanese firms from finan-
cial pressures. Second, liberalization in Japanese
bond markets in the 1980s and again in the early
1990s eliminated regulatory obstacles to the use of
non-bank debt, thereby significantly reducing
traditional dependence on banks for debt financing
(Campbell and Hamao, 1994; Weinstein and Yafea,
1998).3
Despite earlier findings of a lower cost of capital
for Japanese firms (Frankel, 1991), more recent
evidence indicates that the cost of this bank-
centered system may be significant. Weinstein
and Yafeh (1998, 636) note that: 'Before liberal-
ization of Japanese financial markets in the 1980s
firms paid dearly for the privilege of access to bank
finance.' Weinstein and Yafeh (1998) document the
persistence of a significant premium for bank
financing even in the context of recent competitive
pressures. These authors estimate that slightly over
one third of profitability differences between bank-
affiliated and independent firms in their study
could be explained by greater reliance on costly
bank debt incurred by main bank firms. On the one
hand, this increased cost may reflect the premium
paid for the monitoring, support and access to
stable financing made possible by established bank-
ing relationships. On the other hand, increased
transparency in financial reporting and more
stringent regulation may reduce the benefits and
usefulness of the bank monitoring function. The
costs of this system may become increasingly
onerous in the context of global competition.
Indeed, studies have shown sharp increases in the
use of bonds, including foreign bonds, subsequent
Journal of International Business Studies
* lapanese
I-f ---
380
to deregulation. In 1980, firms listed on the Tokyo
Stock Exchange had nearly 90% of their debt in
bank borrowing. This figure had dropped to 50% by
1991 (Hoshi, 1994).4 The impact of this evolution
away from bank financing, however, is unclear
(Kim and Hoskisson, 1996). Gerlach (1992) argues
that these changes become relevant only to the
extent that the behavior and priorities of the
holders of equity and new forms of debt differ
from traditional bank financing. To substitute one
form of financial tie with another form of tie in
such a closely linked homogeneous network may
have little substantive impact. Gerlach (1992,
127-130) and Hoshi (1994) support substitution
of group debt for other forms of group-centered
financing. Indeed, Hwang and Kim (1998) provide
evidence for the substitution of group debt for non-
group debt among keiretsu firms. Further,Roe et al.
(1993) argue that Japan's deeply entrenched system
of powerful intermediation encourages the persis-
tence of strong banking relationships.
Paker and Hodder (2002) observe that, although
both independent and keiretsu firms reduced their
dependence upon bank debt in the 1990s, reduc-
tions in overall leverage and bank debt were more
pronounced for keiretsu firms. Further,Hwang and
Kim (1998) find strategic differences between firms
making greater use of non-bank debt and those
relying more heavily on bank financing Weinstein
and Yafeh (1998) show that the preferential access
to financing available to main bank firms did
not result in higher growth. Indeed, they find
lower financial performance among main bank
firms.
This discussion suggests no clear consensus
regarding the stability of the Japanese inter-corpo-
rate network in the context of economic and
regulatory change. Despite evidence of the weak-
ening of inter-corporate and financial ties (e.g.,
Prowse, 1990), other studies support the persistence
of traditional banking and equity ties (e.g.,
Johnston and Maloney, 1998). There is similar
ambiguity regarding the potential implications of
any such evolution of the Japanese system. Kim and
Hoskisson (1996) and Weinstein and Yafea (1998)
suggest that the costs of Japanese inter-corporate
networks may become more prominent in the
context of increased competition and the globaliza-
tion of financial markets. Gerlach (1992) and
Ramseyer (1994) and Roe et al. (1993) in contrast,
argue that economic changes may have little
impact on the identity or priorities of major
stakeholders. The closely linked Japanese inter-
Journal of International Business Studies
keiretsu
----- ----oraanization
--- ---- I McGuireand S Dow
I.. .--.
corporate network may further limit the ability of
new stakeholders to impose new performance
pressures. This issue has become increasingly con-
troversial in the context of the continued decline in
the Japanese economy, which has challenged the
Japanese model, and of corporate governance fail-
ures such as Enron and Worldcom, which have
challenged the North American model.
In view of this ambiguity, it is difficult to offer
specific hypotheses regarding the persistence of
keiretsu ties and the performance implications of
Japanese industrial organization. As a result, two
research questions are proposed:
(1) To what extent have economic and regulatory
changes influenced the Japanese inter-corpo-
rate network?
(2) To what extent have the performance char-
acteristics of Japanese firms remained stable
over more recent time periods?
Data and methodology
Sample and time frame
Data on keiretsu affiliation were obtained from
IndustrialGroupingsin Japanfor the years 1990-1991
and 1996-1997. This bi-annual reference is the
accepted source of information regarding group
membership and ties and has been used in previous
studies. We use classification as a member of the six
major horizontal keiretsu as listed in Industrial
Groupingsin Japanfor each year to identify keiretsu
firms.5 Thus each sample includes firms identified
as either independent or affiliated in each of the
two time periods. The Japanese sample consists of
434 firms. The US sample of 764 firms was matched
to the Japanese sample for each year in terms of
industry and size. Additional data on inter-firm ties
and ties and control variables are taken from the
Worldscopedatabase for 1992 and 1997.
To assess changes in Japanese industrial organiza-
tion it was necessary to select periods that would
tap changes in the Japanese business environment.
The earlier period was one of boom markets in
Japan.6 Data from the OECDand the International
Monetary Fund (OECD, 1998; Bayoumi, 1999)
show that this period was one of growth and
expansion in the Japanese economy along most
economic indicators. The later period, in contrast,
was one of substantial financial pressures on
Japanese firms and financial institutions. Both
OECD and IMF data suggest 1991 as a pivotal year
 Japanese keiretsu orqanization
in the Japanese economy. Examination of IMFand
OECD statistics shows a clear change in the
strength of the Japanese economy during the later
period.7 The use of these time frames is also
congruent with Geringer et al. (2000), who empiri-
cally identified 1987-1991 and the post-1992
period as distinct strategictimeperiods.
It is, of course, impossible to specify the precise
time frame in which the effects of regulatory
change would be expected. Deregulation in the
Japanese economy has been a gradual process over
the last two decades and into the 21st century.
Several researchersnote that, although the process
of market deregulation had already begun through
a "first wave" in the early 1980s, reforms had not
been fully instituted (Campbell and Hamao, 1994;
Ueda, 1994; Weinstein and Yafeh, 1998). Thus
Paker and Hodder (2002) note differences in firm
capital structure consistent with regulatory change
during the pre-1991 and post-1991 periods. Further,
the impact of early reforms would probably have
been gradual as financial institutions adjusted to
regulatory developments. Further, and perhaps
more significant, reforms were instituted in the
early 1990s, and remained in progress at the end of
the decade (Gibson, 1998).
Measures
Firmclassification
The sample is divided into three mutually exclusive
categories: US, keiretsu, and non-keiretsu firms. In a
sample including three categories, as in this case,
we need only two dummy variables to perform the
analysis (Aiken and West, 1991; Hardy, 1993). As we
are interested in comparing the keiretsu and non-
keiretsu firms with US firms, the latter were
designated as the reference group. Accordingly, for
the first dummy variable, X1, members of the
keiretsu category are assigned scores of 1, and
members of the non-keiretsu category are assigned
scores of 0. For the second dummy variable, X2,
members of the non-keiretsu category are all
assigned scores of 1, and members of the keiretsu
category are assigned scores of 0. In this way, the
regression coefficients will express the difference
between the three group means (Aiken and West,
1991; Hardy, 1993).
Inter-corporatelinkages
We use three indicators of inter-corporate linkages.
To assess equity ties among firms and financial
institutions, we use the percentage of outstanding
j McGuire and S Dow
-38
381
shares owed by institutional investors and those
held in blocks of greater than 5% of outstanding
shares (5%-shareholders). These variables have
been commonly used to assess the strength of
equity ties. Based upon Prowse (1992) and Dow and
McGuire (1999) we anticipate that the "concen-
trated yet dispersed" structure of Japanese owner-
ship will be reflected in lower significant
shareholdings among Japanese firms. Despite argu-
ments for higher institutional ownership in the
Japanese context, Dow and McGuire (1999) observe
lower institutional ownership among Japanese
firms. These findings are theoretically congruent
with the argument of smaller ownership stakes
reinforced by other ties.
Additionally, we evaluate reliance on bank finan-
cing by the ratio of short-term to total debt. Since
bank financing is the primary source of short-term
debt in both the US and Japan, this measure has
been used in previous studies as an indicator of
reliance on bank financing (Rajan, 1992; Campbell
and Hamao, 1994; Dow and McGuire, 1999;
Kanatas and Qi, 2001).
Firmperformance
The previous discussion suggests that Japanese
industrial organization may imply strategic and
performance differences among keiretsu, non-keir-
etsu, and US firms. Specifically, we would expect
lower market and accounting performance among
Japanese firms, particularlykeiretsu firms. Account-
ing performance is measured by 5-year average
ROA. Market performance is measured by the
5-year average price/earnings ratio. The insulation
of Japanese firms from market pressures might also
reduce their market risk. We therefore use firm beta
as an indicator of market risk.
Earlier,we argued that the costs of maintaining
inter-corporate ties might lead to reduced cost-
effectiveness or efficiency among Japanese firms.
Efficiency is measured by the ratio of sales and
general administrative expense as a percentage of
sales.
Growth and foreign involvement
Nakatani (1984), Gerlach (1992) and Weinstein and
Yafeh (1998) also contend that Japanese firms,
particularly keiretsu firms, may pursue more con-
servative growth strategies. Growth is measured by
5-year average sales growth. A similar argument
suggests that Japanese access to a stable domestic
market and more conservative strategic orientation
may imply lower foreign involvement among
Journal of International Business Studies
* Japanese keiretsu oraanization
382
Japanese firms, particularlykeiretsu firms (Hundley
and Jacobson, 1998; Geringer et al., 2000). In
essence, keiretsu firms may tend to make use of
their well-established and protected domestic mar-
ket, whereas non-keiretsu firms may seek growth
through foreign sales. Congruent with Geringer
et al. (2000) we measure activity in foreign markets
by the percentage of foreign sales.8
Controlvariables
We control for firm size (log of net sales) and
industry (two-digit SICcode). AsJapanese firms rely
more heavily on debt than do US firms, we also
control for firm leverage measured using the debt-
equity ratio.
Descriptive statistics for all variables are found in
Appendix A.
Methodology
We use hierarchical regressions to address the two
research questions.9. The rationale underlying the
use of this technique is as follows. The first research
question examines the extent to which Japanese
and US firms have converged in terms of debt and
equity ties throughout the two periods. In parti-
cular, if the economic and regulatory changes in
the early 1990s had their hypothesized effects, it
might be expected that the Japanese firms would
converge on the US model. Our earlier discussion
presented conflicting arguments regarding differ-
ences between keiretsu and non-keiretsu firms. The
greater insulation of keiretsu firms may imply
greater stability. If, however, economic and regula-
tory changes imply pressures on traditional pat-
terns of inter-corporate ties, these pressures may be
especially strong for the keiretsu firms as they have
further to go.10 To investigate the first research
question concerning changes in inter-firm ties we
used three hierarchical regressions for each depen-
dent variable (e.g., institutional ownership, etc.).
The first regression, which investigates the differ-
ences in the means of the three categories on the
dependent variable, includes the set of control
variables as well as the two dummy variables for
firm classification (keiretsu, non-keiretsu). In the
second regression, we added one dummy variable
for time period. Finally, the third regression
incorporates the two interaction terms obtained
through multiplying the two dummy variables for
firm classification and the dummy variable for time
period.
Similarly, the second research question investi-
gates whether the performance characteristics of
Journal of International Business Studies
I McGuireand S Dow
Japanese firms fostered by the Japanese inter-
corporate system remained stable or converged
towards a North American performance profile.
Again, to address this question we used three
hierarchical regressions for each dependent variable
measuring performance (e.g., ROA 5-year average
etc.). The first regression consists of the set of
control variables and the two dummy variables for
firm classification (keiretsu, non-keiretsu). The
second regression adds the dummy variable captur-
ing the time period. Finally, the third regression
includes the two interaction terms obtained
through multiplying the two dummy variables for
firm classification and the dummy variable for time
period firm. We investigate whether there is an
interaction effect in the population, based on the
analysis of sample data, using the following
hierarchical F-test (see Jaccard et al., 1990, 22 and
Hardy, 1993, 24 for F-test details).
(R2 -R2)/(K2 - K1)
FK2-K1,N-K2=1 (1 - R2)/(N - K2 - 1)'
where R2 is the R2 for regression 3, which includes
the control variables, the two dummy variables for
firm classification, one dummy variable for the
time period, and c theterms; R is the
two interaction
R2 for regression 2, which includes the control
variables, the two dummy variables for firm
classification, and the dummy variable for time
period; K2 is the number of predictors in regression
3, and K1 is the number of predictors in regression
2; and N is the total sample size.
Results
To what extent have recent economic and
regulatory changes influenced the Japanese inter-
corporate network?
The empirical results pertaining to research ques-
tion 1 are contained in Table 1. This contains only
the results of full model estimation (regression 3).11
Results obtained in a statistically significant hier-
archical F-test for regression 3 vs regression 2 would
support the existence of an interaction effect.
Results from regression 1, which includes the set
of control variables and the two dummy variables
for firm classification, suggest that both categories
of Japanese firms demonstrate lower institutional
ownership and 5% ownership than do US firms.12
As expected, they also show higher short-term debt
than US firms. Although both categories of
Japanese firms had lower foreign sales than US
firms, this effect is strongest for keiretsu firms.
 Japanese keiretsu orqanization JMcGuire
andS Dow
383

Table 1 To what extent have economic and regulatory changes influenced the Japanese inter-corporate network?

Institutional 5% ownership Shortdebt % Foreign%

ownership total debt total assets

Log net sales 0.509** -0.183** -0.212** 0.427**

(0.239) (0.322) (0.390) (0.186)
Leverage -0.030 -0.007 0.017 0.018
(0.0001) (0.001) (0.001) (0.001)
SICdummy 1 -0.026 -0.017 -0.014 -0.050*
(1.553) (2.094) (2.487) (1.221)
SICdummy 2 0.014 -0.035 -0.053* -0.077**
(0.989) (1.333) (1.604) (0.782)
SICdummy 3 0.010 -0.006 -0.075** -0.017
(2.781) (3.750) (4.420) (2.242)
SICdummy 4 -0.092** 0.056** 0.029 -0.095**
(1.568) (2.114) (2.553) (1.257)
SICdummy 5 -0.083** -0.069** 0.044* -0.135**
(1.574) (2.122) (2.617) (1.193)
SICdummy 6 0.064** 0.036 0.010 -0.069**
(1.305) (1.759) (2.272) (1.000)
Keiretsudummy variable -0.534** -0.249** 0.491 ** -0.391**
(1.307) (1.762) (2.140) (1.067)
Non-keiretsu dummy variable -0.144** -0.21 3** 0.231 ** -0.117**
(2.579) (3.450) (4.231) (2.165)
Yeardummy (1992=1) -0.078** -0.043 -0.006 0.018
(0.956) (1.289) (1.566) (0.724)
Keiretsudummy variable*year 1992 0.104** 0.057 -0.077 -0.107**
(1.6200 (2.184) (2.634) (1.303)
Non-keiretsu dummy variable*year 1992 -0.021 0.057* -0.025 -0.094**
(3.539) (4.751) (5.799) (2.900)
F 47.898** 38.216** 29.619** 44.543**
R2 0.213 0.178 0.146 0.203
R2change (relative to regression 2) 0.005 0.003 0.002 0.008

*P<0.05;**P<0.01.
All P weights are from final step in hierarchicalregression. Standarderrorsare in parentheses.

However, results of the hierarchical F-tests captur-
ing the magnitude of the effects of the interaction
terms were not significant for all estimated models.
These findings suggest that differences in the
patterns of ownership, financing, and inter-corpo-
rate ties between both types of Japanese firm and
US firms have persisted, even after the major
economic and regulatory changes that have taken
place between 1991 and 1996. Our results may
suggest that the evolution and impact of regulatory
change are gradual.
To what extent have the performance
characteristics of Japanese firms remained stable
over recent time periods?
The second research question focuses on the
possible performance shifts over time. The empiri-
cal results of the hierarchical regressions involving
performance measures as dependent variables are
presented in Table 2. Despite the stability of inter-
corporate ties found in the first analysis, changes in
the economic and regulatory context may imply
differing performance pressures. Results from
regression 1 suggest that both categories of Japa-
nese firms show lower performance as measured by
ROA and sales growth. Keiretsu firms have lower
levels of foreign sales, market risk (beta) and sales
expenses than US firms. Results for market perfor-
mance (PIE ratio) are not significant. Addition of
the classificatory variables has little impact on the
R2 of the estimated models. Again, hierarchical F-
tests suggest that there are no significant interac-
tion effects involving firm classification dummy
variables and time period. These results indicate
that the pattern of performance fostered by the
Japanese inter-corporate system remained stable
during the pre- and post-1991 periods considered in
this study.

Journalof InternationalBusinessStudies
* laDanese keiretsu oraanization I McGuireand S Dow
384

Table 2 To what extent have the performancecharacteristicsof Japanesefirms remained stable over more recent time periods?

Dependentvariables
ROA5 years P/E5 years / SG&A% sales Sales growth Foreign%
5 years total sales

Log net sales 0.340** 0.041 0.042 -0.233** 0.146** 0.428**

(0.143) (3.987) (0.011) (1.795) (0.224) (0.233)
Leverage -0.007 -0.014 0.006 -0.006 -0.001 0.007
(0.001) (0.012) (0.001) (0.003) (0.001) (0.001)
SICdummy 1 -0.020 0.021 -0.087 -0.034 -0.001 -0.088**
(0.925) (25.704) (0.074) (11.619) (1.458) (1.544)
SICdummy 2 -0.006 0.002 -0.035 0.016 0.009 -0.156**
(0.590) (16.524) (0.046) (7.415) (0.928) (0.970)
SICdummy 3 0.006 0.018 -0.018 -0.030 -0.006 0.082**
(1.679) (46.582) (0.130) (21.064) (2.610) (2.812)

SICdummy 4 -0.048* -0.022 -0.059 -0.006 -0.010 -0.1 39**

(0.933) (25.950) (0.075) (11.775) (1.4660 (1.568)
SICdummy 5 -0.009 0.007 -0.065 -0.059** -0.075** -0.168**
(0.926) (25.330) (0.075) (11.720) (1.4440 (1.504)
SICdummy 6 0.001 0.003 0.007 -0.023 0.007 -0.072**
(0.765) (21.898) (0.063) (9.615) (1.200) (1.247)
Keiretsudummy variable -0.159** 0.019 0.093** 0.075* -0.258** -0.304**
(0.798) (21.831) (0.062) (10.048) (1.252) (1.320)
Non-keiretsu dummy variable -0.066* 0.185 0.044 0.024 -0.121** -0.067
(1.577) (42.771) (0.120) (19.794) (2.477) (2.622)
Year 1992 dummy variable 0.117** -0.010 0.069** -0.017 -0.017 -0.005
(0.554) (16.142) (0.046) (6.957) (0.870) (0.908)
Keiretsudummy variable* year 1992 -0.033 0.004 0.1 70** 0.006 0.166** -0.027
(0.974) (27.032) (0.076) (12.280) (1.530) (1.617)
Non-keiretsu dummy variable*year 1992 -0.003 -0.101** 0.081** 0.002 0.081 ** -0.035
(2.158) (58.713) (0.165) (27.098) (3.391) (3.540)
F 17.183** 4.122** 18.257** 7.302** 8.170** 39.737
R2 0.82 0.023 0.097 0.037 0.041 0.182
R2change (relative to regression 2) 0 0.005 0.013 0 0.013 0
*P<0.05; **P<0.01.
All standardized B weights are from final step in hierarchicalregression. Standarderrorsare in parentheses.

Conclusions context of Japan, they are consistent with findings

Our results provide little evidence that economic
and regulatory changes in the early 1990s influ-
enced the Japanese inter-corporate network, and in
particular keiretsu organization. In spite of signifi-
cant changes in the Japanese economy during the
early 1990s and the globalization of markets, the
keiretsu system appears to be have remained
intact.13 Perhaps the most striking finding is that,
despite arguments for reduced dependence on bank
financing, we observe little evidence of such
change. In general, our findings do not provide
strong evidence for changes in Japanese industrial
organization or the performance characteristics of
Japanese firms.
Although these findings may appear surprising
given changes in the regulatory and economic
by Gerlach (1992) and Johnston and McAlevey
(1998) of the stability of the inter-corporate net-
work in Japan. This relative stability of inter-
corporate ties would seem to suggest the persistence
of the benefits (and costs) of such ties. One of the
theoretical advantages of the keiretsu system is its
ability to insulate member firms during periods of
crisis or economic downturn in general. On the one
hand, our findings regarding performance support
previous research indicating lower profitability and
more conservative strategic profiles among Japa-
nese firms during both time periods. On the other
hand, we find little evidence suggesting the
insulating effect of keiretsu organization.
We do not find evidence of the evolution of
performance profiles. Although the performance

Journal of International Business Studies

 Japanese keiretsu organization
profile of keiretsu firms differs from that of US firms
to a greater extent than that of non-keiretsu firms,
performance profiles remained stable over the two
periods, exhibiting no tendency to converge to
more North American norms. In general, these
results suggest that, in spite of perhaps subtle
changes in keiretsu ties and perhaps institutional
pressures from global financial markets, we find
little evidence of evolution to a more shareholder
orientation. Further research can build upon stu-
dies that have examined the relationship between
stakeholder pressures and adoption of North
American performance profiles (e.g., Phan and
Yoshikawa, 2000). Analysis can also track the co-
evolution of firm ownership and performance
characteristics.
Sheard (1991) notes that, independent of corpo-
rate governance, other characteristics of Japanese
firms imply reliance on inter-corporate networks.
For example, Japanese firms tend to be relatively
specialized and non-integrated. To the extent that
one function of the financial interlocks examined
in this study may be to solidify and reinforce
operational linkages such as buyer-seller relation-
ships, cooperative research activities, and personnel
exchanges, it may be important to examine the role
of Japanese corporate governance in a broader
context. Bappa (2000), for example, proposes that
inter-corporate networks may represent a valuable
source of social capital to Japanese firms and
their executives. The persistence of networking
patterns may be linked to longer-term and
more deeply imbedded roles for these ties. These
inter-corporate ties, and the social capital they
represent, may be an important source of strategic
options for future development (Bowman and
Hurry, 1993).
This study also suggests several areas for future
research. First,we examined only a limited number
of the type of ties linking Japanese firms. We also
focused on formal ties. An important aspect of any
cooperative network is less formal ties. Further,the
comparative nature of our sample made it difficult
to assess changes in ties more specific to the
Japanese context. Examples of such ties might be
intra-group shareholdings, lending, or purchasing
for which there is no North American equivalent.
Finally, future research can expand the range of
performance variables examined.
Regulatory and economic change is an ongoing
process. Significant regulatory changes occurred
during the 1990s and are still continuing (Gibson,
1998). Recently, the Japan ExternalTradeOrganiza-
J McGuireand S Dow
385
tion (ETRO, 2001) acknowledged that changes in
the Japanese economic context, notably in the
areas of reduced dependence upon bank financing
replaced by direct access to capital markets, reduc-
tion in the level of cross-holding, internationaliza-
tion of management and heightened numbers of
foreign shareholders, have created pressurestoward
a more North American style of corporate govern-
ance. In fact, JETRO(2001) has stated the position
that:
Government and corporations alike have found themselves
undertaking measures to make the basic structure of
corporate governance work, with the structure of authority
flowing from stockholders to board of directors to man-
agers.
A recent article in Japan Times (Fukashima, 2002)
outlines the current governance debate in Japan.
Proponents of Western-style governance argue that
better disclosure, transparency, shareholder rights,
board independence, and reduction in share cross-
holding are among the issues cited as essential to
improving Japan's competitive position. The well-
publicized failure of the US system in the cases of
Enron and Worldcom have undoubtedly cooled
some opinions in this regard. Those who line up
against sweeping reforms of the Japanese govern-
ance system suggest that the US model is incompa-
tible with Japan's socio-economic context. In
particular,they argue that the Japanese corporation
is concerned not just with shareholders, but with a
broader range of stakeholders that include custo-
mers, suppliers, creditors, employees, and commu-
nity. In essence, they are arguing that the Japanese
stakeholders encompass all who are implicated in
the traditional keiretsu, and not simply shareholder
hegemony.
Economic turmoil in Japan has continued. Given
the significant changes and challenges faced in the
Japanese economy during the 1990s and beyond,
the continued success of many Japanese firms may
support the benefits of the Japanese system in
assisting firms during economic change and
decline. The ability of the Japanese system to
continue to weather economic and regulatory
change is uncertain. The answer to this question
remains to be seen. Indeed, the Japanese Commer-
cial Code, the cornerstone of the corporate legal
system, is being revised for the first time in 50 years
with a view to harmonizing the legal and account-
ing systems with its competitors. Thus the time
frame used in this study may, indeed, be only an
intermediate step in this longer process. However,
Journal of International Business Studies
* I ?~ keiretsu
laoanese ~-- oraanization
-I
386
the continued importance of the topic of this
research suggests its ongoing relevance to the field
of international business.
A consistent finding among studies of Japanese
industrial organization is that it is particularly
stable in the context of economic change (Johnston
and McAlevey, 1998; Dow and McGuire, 1999;
McGuire and Dow, 2002). Further research can
explore this future evolution to assess whether the
continuing pressures on the Japanese economy will
eventually lead to fundamental changes in Japa-
nese industrial organization.
Acknowledgements
The authors gratefully acknowledge the financial
support of the Social Sciences and Humanities
Research Council of Canada. Zied Guedri provided
invaluable assistance in this research.
Notes
1This insulation from market pressures is augmented
by Japanese corporate governance practices, which
make it very difficult for outside shareholders to
have any impact on corporate governance (Sheard,
1994a).
21to (1997) further suggests that the industrial
domination of keiretsu companies may limit their
growth in an increasingly service- and technology-
oriented economy. Gerlach (1992), however, argues
that the flexibility and uncertainty absorption offered
by stable shareholdings are particularly critical during
periods of rapid change. He also cites anecdotal
evidence that foreign investors and firms may adopt
more Japanese-style modes of behavior congruent
with the Japanese context.
3lndeed, this trend toward securitization of debt
(replacement of bank financing by other forms of
debt) may also weaken the stable shareholding
arrangements discussed earlier. This logic holds that
reduced reliance on debt makes it more difficult for
banks to balance equity ties with lending. In the
absence of the additional monitoring and input
provided by strong lending relationships, banks may
References
Ahmadjian, C. and Lincoln, J. (2001) 'Keiretsu, governance, and
learning: case studies in change from the Japanese automobile
industry', Organization Science 12(6): 683-701.
Ahmadjian, C. and Robbins, G. (1999) 'Foreign share ownership
and corporate behavior in Japan', Paper presented at Academy
of Management Meetings, Chicago.
Journal of International Business Studies
I McGuireand
-.---. S Dow
hold significant equity positions. Further, greater
liquidity in the security markets may fragment bank
stock holdings (Roe et al, 1993; Gibson, 1998).
4See also Weinstein and Yafeh (1998) and Campbell
and Hamao (1994).
5There is some degree of overlap between horizon-
tal keiretsu and certain vertical (buyer-supplier) net-
works. We eliminated a small number of firms
belonging to independent vertical groupings (those
having no established ties with the six horizontal
groupings).
6Although this period includes the 1987 market
decline, Johnston and McAlevey (1998) and Sheard
(1991) note that the market crash was more limited
and had a short-lived impact on the Japanese market.
Indeed, the period 1988-1989 represented a period of
extraordinary growth on the Nikkei Stock Average
(Johnston and McAlevey, 1998). The IMF (Bayoumi,
1999:3) places the bursting of the asset price bubble in
1991.
7The IMFnotes that, despite some signs of recovery
in 1996, the Japanese economy continued its reces-
sionary tendencies.
8Many of the larger firms in our sample may have
had foreign investment replacing exports. Foreign
assets, however, were less consistently reported,
particularly for the smaller firms in our sample. This
may introduce a bias in our results.
91dentical results were found using a series of chow
tests.
10We thank one of the reviewers for suggesting the
use of hierarchical regressions.
1Results from intermediate steps are available from
the authors.
12For reasons of parsimony we report in Tables 1
and 2 only the results obtained from full model
estimation (regression 3). Intermediate results (regres-
sions 2 and 3) are available from the authors.
3Regulatory change continued to progress
throughout the 1990s and beyond. Our time frame,
however, captures only two distinct strategic periods,
which most likely serves to attenuate the impact of
these changes.
Aiken, L.S. and West, S.G. (1991) Multiple Regression: Testing and
Interpreting Interactions, Sage: Thousand Oaks, CA.
Aoki, M. (1994) 'Monitoring Characteristicsof the Main Bank
System: An Analytical and Developmental View', in M. Aoki
and H. Patrick(eds.) TheJapanese Main BankSystem, Oxford
UniversityPress:Oxford, 109-141.
 ?lananpcpM
JUlU,,
kL.tiratc.
IqIIrCLd
nrnAn;7Ati^n
Vq/IaIIILI/t11
Bappa, A. (2000) 'The Japanese family: an institutionallogic for
Japanese corporate networks and Japanese management',
Academyof management Review25(2): 409-415.
Barber,T. (2002) 'Shareholder democracy gets a bigger say',
FinancialTimes,May 14, 2002, Surveyof GermanIndustry,p. 1.
Berglof, E. and Perotti, E. (1994) 'The governance structure of
the Japanese financial keiretsu', Journalof FinancialEconomics
36(2): 259-285.
Bayoumi, T. (1999) The MorningAfter:Explainingthe Slowdown
in JapaneseGrowthin the 1990s, InternationalMonetary Fund:
Washington, DC.
Bowman, E.H. and Hurry, D. (1993) 'Strategy through the
options lens: an integrated view of resource investments and
the incremental choice process', Academy of Management
Review18: 760-783.
Bruce, R. (2002) 'Rule maker set to change US tradition',
FinancialTimes,March 27, 2002, p. 132.
Campbell, J. and Hamao, Y. (1994) 'Changing Patterns of
Corporate Financing and the Main Bank System in Japan', in
M. Aoki and H. Patrick(eds.) TheJapanese Main BankSystem,
Oxford UniversityPress:Oxford, 325-349.
Caves, R.and Uekasa,M. (1976) IndustrialOrganizationin Japan,
The BrookingsInstitute:Washington, DC.
Corbett, J. (1994) 'An Overview of the Japanese Financial
System', in N. Dimsdale and M. Prevezer (eds.) Capital
Markets and CorporateGovernance,Oxford University Press:
Oxford.
Dewenter, K., Novaes, W. and Pettway, R. (2001) 'Visibility
versus complexity in business groups: evidence from Japanese
keiretsu',Journalof Business74(1): 79-100.
Dow, S. and McGuire,J. (1999) 'The sources and advantages of
Japanese industrial organization', Asia Pacific Journal of
Management 16(1): 47-74.
Flath, D. (1993) 'Shareholding in the keiretsu,Japan'sFinancial
group', The Reviewof EconomicsStatistic75: 249-257.
Frankel, J. (1991) 'The cost of capital in Japan: an update',
BusinessEconomics26: 25-31.
Fukashima, G.S. (2002) 'The corporate governance debate',
Japan Times,April9, 2002.
Geringer, J.M., Tallman, S. and Olsen, D. (2000) 'Product and
international diversification among Japanese multinational
firms', StrategicManagementJournal21(1): 51-80.
Gerlach,M. (1992) AllianceCapitalism:TheSocialOrganizationof
Japanese Business,Universityof CaliforniaPress:Berkeley.
Gerlach, M. and Lincoln,J. (1992) 'The Organizationof Business
Networks in the United States and Japan',in N. Nohiriaand R.
Eccles (eds.) Networksand Organizations,HarvardUniversity
Press:Boston.
Gibson, M. (1998) "Big bang' deregulation and Japanese
corporate governance: a survey of the issues', Board of
Governorsof the FederalReserveSystem InternationalFinance
Discussion Papers, Washington, DC.
Gilson, R.J.and Roe, M.J. (1993) 'Understanding the Japanese
keiretsu: overlaps between corporate governance and indus-
trial organization', YaleLawJournal102(4): 871-907.
Guillen, M. (2000) 'Business groups in emerging economies: a
resource based view', Academyof ManagementJournal43(3):
362-380.
Hardy, M.A. (1993) Regressionwith Dummy Variables,Sage:
Newbury Park,CA.
Harnischfeger, U. (2002) 'Grudging moves on corporate
transparency',FinancialTimes,June 12, 2002, Special Section,
p. 3.
Hoshi,T. (1994) 'The Economic Roleof CorporateGrouping and
the Main Bank System', in M. Aoki and M. Dore (eds.) The
Japanese Firm: Sources of Competitive Advantage, Oxford
UniversityPress:Oxford, 285-309.
Hoshi, T., Kashyap, A. and Scharfstein, D. (1991) 'Corporate
structure, liquidity, and investment: evidence from Japanese
industrialgroups', The QuarterlyJournalof Economics106(1):
33-59.
I AMr,iiir
) Iv1ickUII
... rn
allu
C ..nw
J VuVVW
387
Hundley, G. and Jacobson, C.K.(1998) 'Effectsof the keiretsuon
the export performance of Japanese companies: help or
hindrance', StrategicManagementjournal 19(10): 927-938.
Hwang, L.S.and Kim,Y.O. (1998) 'Does the ownership structure
of debt and equity affect the agency costs of debt? Japanese
evidence. Journalof Accounting,Auditing, and Finance 13(1):
37-67.
Ito, T. (1997) 'Japan's economy needs structural change',
Finance& Development;Washington34(2): 16-19.
Jaccard, J. and Turrisi,R. (1990) InteractionEffectsin Multiple
Regression,Sage: Newbury Park,CA.
JETRO(2001) http://www.jetro.go.jp/it/e/pub/changing2001/
index.html.
Johnston, S. (1995) 'Managerial dominance of Japan's major
corporations', Journalof Management21: 191.
Johnston, S. and McAlevey, L. (1998) 'Stable shareholdings and
Japan's bubble economy: an historical overview', Strategic
ManagementJournal19(11): 1101-1107.
Kanatas, G. and Qi, J. (2001) 'Imperfect competition, agency,
and financing decisions', The Journal of Business 74(2):
307-338.
Kaplan, S.N. and Minton, B.A. (1994) 'Appointments of
outsiders to Japanese boards: determinants and implications
for managers', Journalof FinancialEconomics36(2): 225-259.
Khanna,T. and Palepu, K. (2000) 'The future of business groups
in emerging markets:long-run evidence from Chile', Academy
of Managementjournal43(3): 268-285.
Khanna,T. and Rivkin,J.W.(2001) 'Estimatingthe performance
effects of business groupings in emerging markets', Strategic
ManagementJournal22(1): 45-74.
Kim,H. and Hoskisson,R. (1996) 'JapaneseGovernance Systems:
A CriticalReview',in S.B. Prasad(ed.) Advancesin International
ComparativeManagement,JAIPress:Greenwich, CT.
Major,T. and Lebert, R. (2002) 'A force for change in German
business', FinancialTimes,May 13, 2002, Section 2, p. 20.
McGuire,J. and Dow, S. (2002) 'The Japanese keiretsusystem: a
comparative analysis', Journal of Business Research 55(1):
33-40.
Morck, R. and Nakamura, M. (1999) 'Banks and corporate
control in Japan',Journalof Finance54(1): 319-339.
Nakatani, 0. (1984) 'The Economic Role of FinancialCorporate
Grouping', in M. Aoki (ed.) The EconomicAnalysisof Japanese
Firms,North Holland:Amsterdam, 227-258.
OECD(1998) OECDEconomicOutlook,OECD:Paris.
Parker, A. (2002) 'Conviction gives ammunition to case for
reform', FinancialTimes.,June 17, 2002, Section 2, p. 19.
Paker,B. and James, H. (2002) 'Japanesecapital structureduring
a turbulant period', Social Science Researches Network
ElectronicPaper Series.
Phan, P.H. and Yoshikawa, T. (2000) 'Agency theory and
Japanese corporate governance', Asia Pacific Journal of
Management 17(1): 1-27.
Prowse, S. (1995) 'Corporate governance in an international
perspective: a survey of corporate control mechanisms among
large firms in the US, UK, Japan, and Germany', Financial
Markets,Institutions,and Instruments4(1): 1-63.
Prowse, S.D. (1990) 'Institutional investment patterns and
corporate financial behavior in the United States and Japan',
Journalof FinancialEconomics27: 43-66.
Prowse, S.D. (1992) 'The structure of corporate ownership in
Japan',Journalof Finance47(3): 1121-1140.
Rajan, R. (1992) 'Insiders and outsiders: the choice between
informed and arm's-length debt', Journal of Finance 37(4):
1367-1401.
Ramseyer, J.M. (1994) 'Explicit Reasons for Implicit Contracts:
The Legal Logicto the JapaneseMain BankSystem', in M. Aoki
and H. Patrick(eds. The Japanese Main Bank System, Oxford
UniversityPress:Oxford, 231-254.
Roe, M.J., Ramseyer, J.M. and Romano, R. (1993) 'Some
differences in corporate structure in Germany, Japan, and
the United States', YaleLawJournal 102(8): 1927-1950.
Journal of International Business Studies
*- Jlaanese keiretsu oraanization I McGuire and S Dow
388

Rossant, J., Ewing, J. and Bremmer, B. (2002) 'The corporate
cleanup goes global', BusinessWeek3781: 80-81.
Sheard, P. (1991) 'The economics of Japanese corporate
organizationsand the structuralimpediments debate: a critical
review', Japanese EconomicStudies 19: 30-78.
Sheard, P. (1994a) 'Bank executives on Japanese corporate
boards', Bank of Japan Monetary and EconomicStudies 12:
2-67.
Sheard, P. (1994b) 'Reciprocal delegated monitoring in the
Japanesebanking system', Journalof Japaneseand International
Economics8: 1-21.
Sheard, P. (1994c) 'The Governance of FinancialDistress',in M.
Aoki and H. Patrick(eds.) The Japanese Main Bank System,
Oxford UniversityPress:Oxford, 188-230.
Sheard, P. (1994d) 'Interlocking Shareholdings and Corporate
Governance', in M. Aokiand M. Dore (eds.) TheJapaneseFirm:
The Sources of CompetitiveStrength, Oxford University Press:
Oxford, 310-349.
Steers, R., Shin, Y.K. and Ungson, G. (1989) The Chaebol,
Ballinger/Harper& Row: New York.
Stewart, S. (2002) 'OSC calls for US-style standards', Financial
Post, Toronto, August 20, 2002, A1.
Suzuki, S. and Wright, R. (1985) 'Financial structure and
bankruptcyriskin Japanesecompanies', Journalof International
BusinessStudies 16(1): 97-110.
Thomas, L.G. and Waring, G. (1999) 'Competing capitalisms:
capital investment in American, German, and Japanesefirms',
StrategicManagementJournal20(8): 729-748.
Ueda, K.(1994) 'Institutionaland RegulatoryFrameworksfor the
Main Bank System', in M. Aoki and H. Patrick (eds.) The
Japanese Main Bank System, Oxford UniversityPress: Oxford,
81-108.
Useem, M. (1998) 'Corporate leadership in a globalizing equity
market', Academyof Management Executive12: 43-59.
Weinstein, D. and Yafeh, V. (1998) 'On the costs of a bank-
centered financial system: evidence from the changing main
bank relations in Japan',Journalof Finance53: 635-672.
Williamson, O.E. (1991 a) 'Comparativeeconomic organization:
the analysis of discrete structuralalternatives', Administrative
Science Quarterly36(3): 269-296.
Williamson, O.E. (1991b) 'Strategizing, economizing, and
economic organizations', Strategic Management Journal
12(Special Issue): 75-94.
Yoshikawa,T. and Gedajlovic, E. (2002) 'The impact of global
capital market exposure and stable ownership on investor
relations practices and performance of Japanese firms', Asia
PacificJournalof Management 19(4): 525-540.
Yoshikawa, T. and Phan, P. (2001 a) 'Alternative corporate
governance systems in Japanese firms: implications for a shift
to stockholder-centered corporate governance', Asia Pacific
Journalof Management 18: 183-205.
Yoshikawa,T. and Phan, P. (2001 b) 'The impact of ownership
structure on Japanese corporate governance', Unpublished
manuscript.
Appendix A Means, standard deviations and
correlation matrix
See Table Al.

Table A1

Variable Mean s.d. 1 2 3 4 5 6 7 8 9 10 11

1 Sales 12.53 2.04 1.00

2 Leverage 131.03 813.54 0.1* 1.00
3 Institutional 29.75 19.73 0.153* -0.037 1.00
ownership
4 5% ownership 30.36 26.04 -0.32* -0.05* 0.066* 1.00
5 Short-term 41.70 31.06 0.065* 0.045* -0.16* -0.09* 1.00
debt
6 ROA 2.11 11.17 0.215* 0.006 0.15* -0.004 -0.06* 1.00
7 PIEratio 22.30 297.57 0.066* -0.001 0.029 0.03 0.017 0.021 1.00
8 p 1.01 0.85 0.152* 0.034 0.053* -0.13* 0.092* -0.038 0.003 1.00
9 Efficiency 25.90 136.86 -0.16* -0.024 0.039 0.04* -0.011 -0.313 -0.005 -0.004 1.00
10 Sales growth 5.15 17.27 0.043* -0.003 0.125* 0.008 0.10* -0.04* 0.002 0.079* -0.05* 1.00
11 Foreign sales 11.57 19.06 0.215* 0.01 0.229* -0.036 0.04* 0.049* 0.014 0.053* -0.009 0.041* 1.00
*Coefficients
aresignificantat the 0.01 level(2-tailed).

Acceptedby Tom Brewer,outgoingEditor,23 February2003.

Journalof InternationalBusinessStudies