Content uploaded by Sanford M Jacoby
Author content
All content in this area was uploaded by Sanford M Jacoby on Jan 09, 2019
Content may be subject to copyright.
Social Science Japan Journal Vol. 8, No. 1, pp 43–67 2005 doi:10.1093/ssjj/jyi012
Published online February 3, 2005
© The Author 2005. Published by Oxford University Press on behalf of Institute of Social Science, University of Tokyo. All rights reserved.
Corporate Organization in Japan and the United
States: Is There Evidence of Convergence?
SANFORD M. JACOBY, EMILY M. NASON AND SAGUCHI KAZURO
¯*
We investigate the changing structure of Japanese and US companies and ask whether there are signs of
national convergence in corporate organization. We present three types of evidence to address this question:
longitudinal data, cross-sectional survey data and structural equation models (SEM). The models are ideal
types of Japanese and US companies and relate human-resource strategy and corporate governance to
organizational outcomes such as employment practices and the role of the executive human resources (HR)
function. We find mixed evidence of convergence.The longitudinal data show some Japanese companies
becoming more like those in the US, and the SEM results show a Japanese-style model emerging in some US
companies. However, there is also evidence of continuing differences in corporate governance, employment
and executive-decision-making in Japan and the US.
1. Introduction
Globalization of product and financial markets is causing inquiry into the possible convergence of
previously distinct national institutions, everything from regulatory systems to corporate governance
to employee relations (Hall and Soskice 2001). Yet much of the literature on the homogenization
of the ‘varieties of capitalism’ has a speculative quality to it. There are few empirical studies, and
those that exist tend to focus on macro-level institutions. The findings are equivocal, with some
studies supporting the occurrence of convergence and others showing little evidence of it (Dore
2000; Guillen 2001; Katz and Darbishire 2000). There is also disagreement in the literature on
what, precisely, the term convergence means. For some, it implies a unidirectional movement
Sanford M. JACOBY is the Howard Noble Professor of Management, History, and Public Affairs at UCLA. He studies
comparative political economy, labor markets and management history. His most recent book is The Embedded Corporation:
Corporate Governance and Employment Relations in Japan and the United States (Princeton University Press, 2004). You may
contact him at sjacoby@anderson.ucla.edu.
Emily M. NASON is a Ph.D. candidate in Management at the UCLA Anderson School of Management. Her research
interests are the role of intuition and ambiguity in fairness judgments, negotiation behaviors and cross-cultural studies of
human resource and knowledge management. She can be contacted by e-mail at emily.nason@anderson.ucla.edu.
SAGUCHI Kazurois a Professor in the Graduate School of Economics at the University of Tokyo, specializing in
employment relations. He is the co-editor, with Hashimoto Shuichi, of Jinji Romu Kanri no Rekishi Bunseki (The History of
Human Resource Management in Japan, Minerva Shobo, 2003). His current research focuses on the relationship between
employment systems and welfare systems in Japan. He can be contacted by e-mail at saguchi@e.u-tokyo.ac.jp.
*This research was supported by the Center for Global Partnership’s Abe Fellowship program, the University of Tokyo’s
Center for International Research on the Japanese Economy (CIRJE), Doshisha University’s Institute for Technology,
Enterprise, and Competitiveness (ITEC), the U.C. Institute of Labor & Employment, the Anderson School’s Center for
International Business Education and Research, and the UCLA Academic Senate. Gerhard Helleman of UCLA’s Psychology
Department played a vital role in the analyses of the primary data. The authors are grateful to the generosity of many other
persons who made this project possible, including Araki Takashi, Chris Erickson, Eve Fielder, Andrew Gordon, Inagami
Takeshi, ItoKen’ichi, Minatani Keiju, Nakamura Keisuke, Nakata Yoshifumi, Nitta Michio, SatoHiroshi, Sekiguchi Tei’ichi,
Suzuki Fujikazu, Suzuki Yoshiji, Takada Kazuo, Takahashi Masayasu, Takata Satoshi, Wakumoto Yoshihiko and Lai-Yong
Wong.
44 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
towards Anglo-American values, institutions and practices. For others, it means conformity to an
optimally efficient pattern—such as privatization or decentralization—that transcends the institu-
tions of any particular nation. Still others collapse these distinctions and contend that the Anglo-
American pattern is optimal (Cronin 2000).
At the enterprise level, the convergence debate has focused on two issues: corporate governance
and employment relations. Corporate governance comprises the rules, practices and procedures by
which managers are held accountable to those who have a legitimate stake in the enterprise, as well
as the definition of who those legitimate stakeholders are. Issues of corporate governance include
the duties of directors, methods of corporate financing, executive compensation, acquisitions and
divestments and other strategic decisions. Whether employees should have a role in corporate
governance is a vital issue, both in Japan—whose postwar system of corporate governance is under
challenge—and in the US, where corporate governance is being reconsidered after the scandals that
occurred at Enron, WorldCom and other firms. Giving employees a substantial voice in corporate
governance is associated with a ‘stakeholder’ model of corporate governance, whereas privileging
stockholders we might term the ‘shareholder’ model (Jacoby 2004).
Employment relations include the presence or absence of internal labor markets that give insiders
preference when filling vacancies, minimize lay-offs during business downturns, and pay attention
to equity in pay setting. Conversely, some companies lack internal labor markets and give little or
no preference to incumbents, are quick to discharge during downturns, and rely on market criteria
in setting pay. Employment relations also include the structure of management decision-making for
employment. Internal labor markets are associated with centralization of decision-making, that is,
with a strong human resources (HR) department at corporate headquarters. Conversely, a market-
oriented approach is associated with decentralized management and a relatively weak HR function.
Recent research demonstrates a link between corporate governance and employment relations.
When governance privileges shareholders, companies find it difficult to sustain firm-specific invest-
ments in human capital and are inclined to shift adjustment risks (e.g. during a business downturn)
to employees. Conversely, a stakeholder approach is consistent with firm-specific human capital
accumulation and the sharing of risks—and rents—between shareholders and other stakeholders,
such as employees (Tachibanaki and Taki 2000).
Another factor of importance is a company’s strategy for utilizing human resources, which is
related to its business strategy. Efficiency wage theory offers a way of conceptualizing HR strategy.
It starts from the observation that there is dispersion of occupational wage rates within industries,
with firms paying above and below the median, even after controlling for labor-quality differences.
Companies paying above the median are said to be following an efficiency-wage strategy, in which
premium wages raise the cost to employees of dismissal, thereby boosting employee effort and
reducing monitoring costs. Efficiency wages also reduce turnover and make it economically feasible
for companies to train employees. These human-capital investments transform employees into quasi-
assets. Conversely, companies paying below the median experience higher turnover and are much
less likely to invest in employee skills. Here, employees are viewed not as assets but as commodities.
Companies accommodate high turnover by continually seeking to reduce their labor costs (Groshen
1991; Levine et al., 2003).
The decision of whether or not to pay efficiency wages is usually determined by, and articulated
with, a company’s overarching business strategy (Bird and Beechler 1995). Viewing employees as
assets derives from a resource-based view (RBV) of business strategy, in which companies seek
competitive advantage from inimitable resources that other companies do not possess, such as
intellectual property or unique human and physical assets (Barney 1991). Resource-based strategy
is inward looking and, as such, concerned with strengthening organizational processes that make a
company distinctive. Employees—human capital—are a key resource. Hence, an RBV based on
employees should be associated with an efficiency-wage approach to HR strategy. Conversely, com-
panies paying below-median wages—for whom labor is simply a cost to be minimized—are likely to
seek competitive advantage in factors other than labor, such as technology or entry barriers.
We present three examples drawn from recent experience. Wal-Mart, a multinational discount
retailer, derives competitive advantage from its inventory system and supplier relationships, which
keep its costs and prices low. It pays at the lower half of the compensation scale and experiences high
turnover of its non-union workforce. Less well known is Costco, another major discount retailer that
competes against Wal-Mart. Costco relies on its trained and loyal employees as a key component of
its business strategy, which combines low prices with good service. It pays its unionized workers
above-median wages and has lower turnover as well as higher labor productivity than Wal-Mart
(Holmes and Zellner 2004).
In Japanese retailing, a similar comparison is that between Ozeki and Seiyu, both supermarket
chains. Ozeki tries to minimize its reliance on part-time employees because, according to its CEO,
full-time (and better paid) employees are more sensitive to customer needs, which boosts customer
retention. Over 70% of Ozeki’s employees work on a full-time basis, whereas at Seiyu, over 70% of
employees are part-timers. Wal-Mart acquired a major stake in Seiyu in 2002 and five Wal-Mart
executives now sit on Seiyu’s board. Seiyu is adopting Wal-Mart’s shipping and display methods as
well as its approach to distribution and supplier relationships (Troy 2003).1
A third pair of contrasting companies comes from the US fast-food industry: McDonald’s and a
competitor called Chick-fil-A. McDonald’s derives competitive advantage from key geographic
locations (a kind of entry barrier), from its production technology, and from frequent price
promotions. Like Wal-Mart, it pays at the low end of the scale and has very high turnover. By
contrast, Chick-fil-A rarely uses price promotions, yet it has very strong customer-retention rates. It
pays 30–50% more than competitors and has a lower turnover of store managers and employees.
High wages translate into reduced supervision costs (it has only 25 people on its field staff) and a
committed workforce that regularly develops new ideas for product improvement (whereas
McDonald’s management exclusively handles product and process improvements.) Unlike Costco,
however, Chick-fil-A is a privately held company, a relevant factor in the strategy–governance
relationship (Reichheld 1996).
Human resources strategy and business strategy are not usually considered in the literature on
convergence, although a few scholars have recognized that particular approaches to strategy are
dominant in particular countries and that this may be related to nation-specific institutions, such as
innovation systems and regulatory policies (Mowery and Nelson 1999). Hence, it is our inference
that national differences in corporate governance may be related to differences in corporate strategy.
During the 1980s, the relationships between these variables—corporate governance, strategy and
employment—displayed different patterns in large Japanese and US companies. The central Japanese
tendency—the ideal-type ‘J-firm’—was a company where corporate governance was organized along
stakeholder lines, with executives balancing the interests of different constituencies (Aoki 1988).
Business strategy was resource-based, which resulted in a powerful headquarters HR function and
in HR strategies that emphasized employees as assets. Employment relations were organized around
internal labor markets featuring long-term employment and rewards tilted to organizational criteria
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 45
1. Thanks to Oh Hak-Soo of the Japan Institute of Labor for information on the retail industry.
such as equity and seniority. Associated with the J-firm pattern was a strong central HR department,
which acted as a balancing wheel between governance, strategy and employment. By contrast, in the
US during the 1980s, the typical pattern—the ‘US firm’—had shareholder-oriented corporate
governance and market-oriented employment policies. Human resource units at headquarters
tended to be relatively weak and overshadowed by the finance function.
What happened to these national patterns as globalization intensified in the 1990s? According to
some observers, the J-firm came under pressure from foreign investors and from economic problems
in Japan, causing a shift in corporate governance, strategy and employment relations towards a US
pattern (Katz 2003). In the US, there was an intensification of an existing tendency to favor
shareholders as a result of new ownership patterns, changing social norms and weaker unions. One
result was for the average large US corporation to tilt further towards a market orientation in HR
strategies and employment relations (Cappelli 1999).
An alternative view, however, is that change was slow in Japan during the 1990s and that the J-
firm constellation of variables remained prevalent (Inagami and Research Institute for the
Advancement of Living Standards 2000). And an alternative view of the US is that, although some
companies intensified their shareholder and market orientation, others began to adopt an RBV
approach to business strategy, an efficiency-wage HR strategy and a mild stakeholder ethos (Pfeffer
1998). Although no longer fashionable to say so, some of these ideas and institutions had Japanese
origins, thus suggesting the occurrence of bidirectional convergence.
Our research compares the present arrangement of corporate governance, strategy and employ-
ment relations in large Japanese and US companies. We rely on a unique data set based on surveys
of senior HR executives of large public companies in Japan and the US. The surveys cover a wide
range of organizational and attitudinal issues and were carefully designed for cross-country
comparability.
We present three types of evidence concerning convergence. First, we have longitudinal data on
changes over the past five years in surveyed companies and changes in the values of Japanese
executives based on the replication of a survey done in Japan in 1993. Secondly, we examine cross-
sectional variable means to determine whether there are statistically significant differences between
Japan and the US. While this does not directly test for convergence, the assumption is that we would
not see statistically significant differences if convergence were extensive. Thirdly, we use structural
equation modeling (SEM) to compare groups of variables and the relationships between them.
Unlike the other two analyses, which compare individual variable levels, these models examine
multiple variable relationships or ‘structure.’ The SEM analyses can tell us whether the J-firm
pattern persists in Japan concurrently with an emerging US model and, for the US, whether the
idealized US-firm model exists and also whether there is evidence of an emerging J-firm model.
Again, while the SEM analyses are based on cross-sectional data, the inference is that the presence
(or absence) of a US model in Japan or of a Japanese model in the US would be consistent with the
presence (or absence) of convergence.
To preview our results, we find evidence both to support and to challenge the notion of
business–system convergence. The longitudinal data show Japan and the US moving in the same
direction, although at different rates. Perhaps because of this difference in change rates, national
means remain significantly different. Finally, while the SEM analyses show some similarities in
structural relationships between Japan and the US, the US model is not strongly identified in
Japan.
46 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
2. The Japanese Corporation in the 1980s
Japanese corporate governance in the postwar decades was a form of stakeholder capitalism in
which, at least until recently, ‘nobody gives a great deal of thought to owners. Firms are not seen as
anybody’s “property.” They are organizations—bureaucracies much like public bureaucracies that
people join for careers, become members of. They are more like communities’ (Dore 2000: 25) The
community was run by its board of directors; boards were larger than those of US and British firms,
some of them having 30–40 (or more) members, although usually a subgroup, the management
committee headed by the president, made key strategic decisions. The board of directors was
composed of incumbent managers serving staggered terms of two to six years, with at least one
director from each major functional department at headquarters as well as from major divisions.
One view of Japanese corporations in these years is that they lacked definable business strategies
and pursued growth and market share by trying to ser ve all market segments. A different view,
however, is that companies did, in fact, have distinctive strategies based on cultivating unique
internal competencies and resources. They competed on the basis of these inimitable assets, the
existence of which was tied to HR strategies based on efficiency wages and firm-specific human
capital (Itami 1987). Corporate governance thus supported a resource-based approach to business
and HR strategies, and these in turn affected employment practices, as shown in Figure 1.
To oversee their HR activities, large Japanese companies had sizeable headquarters HR units,
more than twice as large as their US counterparts and with a reputation for being powerful (Inohara
1990: 7). These units managed an employee’s ‘lifetime’ career, creating an association between
internal labor markets and centralization (see Figure 1). Headquarters also consulted with enterprise
unions and presented labor’s views to the board of directors via the director representing the HR
function. The HR director had the reputation of being a ‘king-maker’ because of his control over
managerial rotations and the selection of senior executives (Itoh 1994). An oft-cited symbol of
power were the performance dossiers HR kept on every employee in the company. Although
decisions on who would be promoted to the top strata of the company—including board mem-
bership—were made by the company president, typically the president consulted the HR director.
Thus HR’s king-making role was related to corporate governance: By socializing managers to a
company perspective and selecting the best managers to govern the community, the HR unit insured
that the company would be in the hands of those who were competent and had shown themselves
to possess traits such as honesty and fairness, necessary for a fiduciary role.
The power of HR also stemmed from the strategies and structure of Japanese corporations. In the
1980s, over three-fourths of large US companies had adopted the decentralized M-form structure,
but many large Japanese companies (around 50%) still had a functional or U-form structure in which
sales, purchasing, accounting, planning, HR and other staff responsibilities are centralized at
headquarters (Fruin 1994: 220). Even companies with the M-form structure had fewer divisions
and less unrelated diversification than comparable M-form companies in the US (Kagono et al.
1985: 40). The role of headquarters was to achieve synergy across units of the company, especially
through the mobilization of intangible corporate resources. This led to HR strategies based on
highly trained and centrally allocated employees (Kono 1999).
The same factors that boosted HR’s status also made for relatively weak finance departments.
Because corporate divisions were closely related in terms of technology and markets, and because of
long-term employment and managerial rotation, senior Japanese executives tended to be well
rounded and did not depend heavily on financial criteria for decision-making (Imai and Itami
1984).
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 47
48 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
Thus, we can model the large Japanese corporation of the 1980s as shown in Figure 1. There was
a positive association between HR strategy and corporate governance, which influenced employ-
ment practices and the centralization of the HR function. Stakeholder corporate governance was
positively associated with the HR function possessing influence over executive decisions and power
compared with other corporate functions. Also, we hypothesize that the impact of governance and
strategy on employment outcomes was mediated by the values held by the company’s executives:
whether employee-oriented values (e.g. concerned about employee morale) or finance-oriented
(e.g. concerned about share price). Alignment between HR strategy, stakeholder governance and
values meant that senior executives with employee-oriented values were associated with strong
internal markets and a centralized approach to managing employment. Alignment developed as a
result of internal socialization of executives, including previous membership in enterprise unions
(Gordon 1998). Finally, we hypothesize that there was a direct, positive relationship between
internal labor markets and HR centralization.
Figure 1. Theoretical Model.
Note: Variables and links in solid lines represent ideal-type Japanese firms (stakeholder model), those in broken
lines represent ideal-type US firms (shareholder model) and those in bold are present in both models.
3. The US Corporation in the 1980s
During the 1950s and 1960s, large US companies professed a mild stakeholder approach to
corporate governance. However, the approach began to break down in the 1970s as shareholders—
notably institutional investors—became more assertive of their rights. Corporate law had always
insisted that boards had a fiduciary responsibility only to shareholders, and shareholders increasingly
pressed boards to restructure companies and squeeze more ‘value’ out of them (Blair 1995). A
‘market for corporate control’ emerged in the 1980s, as evidenced by a wave of mergers and hostile
acquisitions; this activity was facilitated by the development of new financial instruments such as
junk bonds.
Business strategy in US companies was typically focused on external strategies such as product
differentiation and barriers to entry, with little attention to HR strategy other than a concern to
minimize labor costs and dependence on employees. Employee turnover rates in the 1950s and
1960s were much higher than in Europe or Japan. As pressure from shareholders increased in the
1980s, companies increasingly began to focus on reducing labor costs. With companies taking on
more risk to satisfy shareholder demands for higher returns, they transferred that risk to employees.
There were waves of lay-offs in the 1980s and early 1990s, even in companies that were relatively
profitable (Kletzer 1998).
Throughout the postwar decades, HR was regarded as a relatively weak function. Human resources
executives tended to be the ‘low man’—and not infrequently woman (a signal of the function’s low
status)—in the corporate hierarchy. Human resources executives were accused of being too ‘soft’
and lacking business acumen (Ritzer and Trice 1969).
By the 1960s, the M-form model, which facilitated corporate diversification and decentralization,
was well established in the US (Chandler 1962). The M-form brought to ascendance the corporate
finance function, which increasingly came into conflict with HR. A substantial number of CEOs
now came out of finance instead of production. Financial criteria grew more important for deter-
mining the internal allocation of capital and the acquisition of business units (Fligstein 1987).
Lacking quantitative indicators to demonstrate their contributions, HR executives reported that
they had more conflict with finance departments than any other functional unit. From HR’s
perspective, the problem was that ‘[i]n top management about the only thing that counts is finance’
(McFarland 1962: 63).
HR executives sought to boost their status in the US through professionalization, including
professional organizations, certification and the like. Also, the spread of the behavioral sciences—in
selection, attitude surveys, executive development and other areas—helped to raise HR’s status by
linking it to university-based research (Rush 1969). Finally, familiarity with the details of employ-
ment laws—which proliferated in the 1960s and 1970s—helped to legitimize the HR function (Dobbin
and Sutton 1998). But things went from bad to worse for HR in the 1980s as government’s role
shrank and corporate governance became more focused on shareholders. In addition to downsizing,
companies decentralized operations so as to put business units closer to the market and reduce their
dependence on corporate headquarters. Headquarters HR departments found themselves a primary
target of efforts to outsource and get rid of ‘bureaucracy’. The net result was a shrinkage in HR
staffs and in the ratio of HR staff to employees (Kramer 1999).
We can model the US corporation of the 1980s as shown in Figure 1: Shareholder-oriented
governance and a market-oriented HR strategy were positively associated with each other and were
negatively related to internal labor markets and to HR centralization. Also, shareholder corporate
governance was negatively associated with the HR function’s having influence and power. Again, we
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 49
hypothesize that the impact of governance and strategy on employment outcomes was mediated by
the values held by the company’s executives. There was a positive alignment between market-
oriented HR strategy, shareholder governance and finance-oriented values. Top executives aligned
their values to those considered important by the board of directors (Agle, Mitchell, and Sonnenfeld
1999). Senior executives with finance-oriented values rejected internalized employment and a
centralized approach to managing employment. Companies tended to select senior executives with
congruent values, because these individuals were more committed to executing strategy effectively.
Also, we hypothesize that there was a positive relationship between internal labor markets and HR
centralization.
4. Changes in the 1990s
4.1 Japan
Corporate governance is changing in Japan, partly in response to domestic pressures—such as the
unwinding of cross-shareholding—and partly in response to demands from foreign investors that
Japanese companies adopt shareholder-oriented practices such as independent corporate boards. In
the late 1990s, Sony, a bellwether company, created the so-called corporate officer system, whereby
it substantially shrank its board of directors while adding outside directors to it. In many companies
with the Sony system, the HR executive no longer serves on the board and the ‘insider’ perspective
is being diluted by outsiders. At the same time there have been numerous legal reforms to facilitate
a shareholder-oriented approach to corporate governance (Araki 2000). However, there is also
evidence that corporate governance practices are changing at a rather sluggish pace (Ahmadjian
2002).
Factors that once gave Japan a distinctive approach to employment also are eroding. In the area
of HR strategy, Japanese companies are becoming more commodity-oriented, as evidenced by
‘hollowing out’ (the transfer of production jobs to Asia), lay-offs and pay cuts and heavier reliance
on temporary and part-time employees. Budgets for employee training and welfare expenditures,
previously a justification for HR centralization, are being cut. Finance-oriented values are on the
ascendant and with this are coming more market-oriented employment practices (Abe 2002). What
remains unclear is the degree to which the ‘lifetime’ employment system is being dismantled or
whether recent changes represent an attempt to preserve a shrinking ‘core’ of permanent employees
by subsidizing its cost with more transient and lower-paid labor (Kato 2001). Also unclear is
whether changes in HR strategy are being driven by new approaches to business strategy, such as a
shift in relative emphasis from product quality to price, and how these changes are affecting the
values of senior executives. It could be that executives are experiencing role conflict, such that their
community-oriented values are lagging behind the emergence of a new finance-driven model.
Alternatively, values might be realigning to fit new governance and strategy patterns.
4.2 United States
Corporate governance in large US corporations continues on the trajectory established in the 1980s:
an ever stronger emphasis on shareholder concerns. As evidence of this, consider stock options,
which are regarded as a way to align managers with shareholders. While fewer than a third of CEOs
received options in 1980, 15 years later nearly all did. During that period, the share of CEO pay
deriving from options rose steadily, eventually overtaking base salary (Cassidy 2002; Bebchuk, Fried,
and Walker 2002).
50 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
Employment at large corporations became ever more unstable and market-oriented in the 1990s.
This development was driven by a shift to a commodity-type HR strategy at many US companies.
As for HR executives, they struggled with their values and organizational role. Some rejected
operational responsibilities in favor of focusing on executive and strategic issues. This required either
outsourcing HR activities or giving greater responsibilities to line managers. The result was a smaller
headquarters HR unit and the adoption of a ‘business partner’ role, which means no longer
‘pacifying disgruntled employees [but] consulting with internal customers’ and aligning with the
dominant finance-oriented ethos (Csoka 1995: 31).
Yet there are also signs of a countervailing movement at some other US companies: the adoption
of a resource-based view (RBV) of business strategy in which companies focus on their core
competence. For HR strategy, this means a shift to viewing employees as assets instead of com-
modities (Wright et al. 2001). Data from the US automotive industry confirm a shift from a ‘Fordist’
system to a resource-based approach (Lieberman and Dhawan 2001). In other industries too, there
is evidence that firms are utilizing innovations such as employee teams, quality circles, job rotation,
intensive training and other so-called ‘high performance work practices’ (Applebaum, Bailey, Berg,
and Kalleberg 2000). In the service and technology sectors, there is explicit recognition of the
contribution of intellectual capital and training to competitive success. In many instances, the shift
can be traced to Japanese ideas on strategy and organization (Cole 1999).
In RBV-type companies, HR executives emphasize the need to develop and conserve intellectual
capital, to develop distinctive organizational cultures and to foster high-commitment employment
relations. New conceptions of business and HR strategy mean that employees—and the HR func-
tion itself—can be construed not as cost burdens but as sources of competitive advantage. In theory,
at least, headquarters HR then becomes responsible for creating a corporate culture that encourages
employee commitment and for developing practices to support competitiveness, as in the link
between employee retention and customer satisfaction (Ulrich 1997).
Thus, there appears to be a realignment of strategy on the one hand and employment and the HR
executive role, on the other, such that some US corporations which previously followed the US
model are adopting the Japanese model shown, even including modest steps towards stakeholder
governance. Conversely, there is also corroborating evidence indicating that the traditional US
model shown in Figure 1 continues to exist in many US companies.
There is, however, at least one structural difference between Japan and the US, which has to do
with the organization of top management. In large Japanese companies, there is a standard way of
organizing the HR function with respect to the reporting relationships of its senior executives. But
in the US, reporting relationships have been changing. According to our 2001 data, 65% of senior
HR executives in large US companies now report to the CEO, which is a major change since 1977,
when only 30% of senior HR executives reported to the CEO (Janger 1977: 37). There are alternative
interpretations of this development. On the one hand, CEO reporting could be consistent with a
shareholder-oriented ethos and a commodification of employment because, in the same period that
CEO reporting became more prevalent, the role of the CEO also changed. CEOs increasingly were
outsiders selected by company boards for their allegiance to shareholders and their willingness to
slash payrolls (Khurana 2002). Alternatively, CEO reporting is consistent with a drift towards a
stakeholder approach, in which HR executives are on the same level as CFOs and able to
challenge a purely financial approach to management (Price and Walker 1999). We do not take a
position on this issue. Rather, we hypothesize that, whichever view is correct, CEO reporting is
likely to mean a boost in the power and strategic influence of the headquarters HR unit, as shown
in Figure 1. We shall return to Figure 1 in our SEM analyses. It represents the baseline model
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 51
for determining whether either, neither or both ideal-type business systems can be identified in each
country.
5. Survey Design and Data Collection
5.1 Sample and Survey Characteristics
The data used in this study were collected in 2001 from a mail survey of senior HR executives in
large public Japanese and US companies. The sampling frame for Japan consisted of 1008 firms
listed on major Japanese stock exchanges for whom the name of the senior HR executive was
available from a commercial database. For the US, it consisted of 977 companies listed on the New
York Stock Exchange for whom the names of both the senior human resource executive and the
senior finance executive were available in a different database. We specified that the survey be
answered by the top executive in the headquarters HR unit.
We had usable responses from 229 Japanese companies and from 145 US companies, representing
a response rate of 22.7% and 14.5%, respectively.2While our response rate may seem modest, keep
in mind that élite surveys—in this case of senior corporate executives—typically have comparable or
lower response rates. We conducted two rounds of mailings, supplemented by follow-up telephone
calls to non-respondents. While there is the possibility of response bias, we found no difference in
the industry distribution of the respondents and non-respondents. For ease of interpretation, we
decided against using an imputation algorithm for the missing data and, instead, used a reduced
Japan (US) dataset of 181 (105) observations without missing data.
Measured by employment, the mean (median) firm size in Japan was 5083 (2215); in the US it
was 18,259 (5200). The dominant 1-digit sector was manufacturing, which provided 41% of
respondents in the US vs. 59% in Japan. The average age of the respondents in Japan was 52 vs. 48
in the US. None of the Japanese respondents was female vs. 33% in the US.
It should be noted that, when the surveys were conducted in the first half of 2001, each country
was at a different stage of the business cycle. The US was at the tail end of a boom, with low
unemployment and lingering concerns about labor shortages, while Japan was entering its second
‘lost decade’, during which employment and profits grew slowly or, in some instances, contracted.
Because all the data on a particular company come from the same respondent, there is the risk of
common method variance, which can cause biased results. However, bias is minimized because our
executives are reporting, in part, on objective facts (e.g. hiring policy). Our data set is cross-sectional
but it contains several items that ask executives to gauge changes during the previous five years. In
addition, we replicate items from a 1993 study of executive values in Japan, thus giving an eight-
year frame for analyzing change.
5.2 Variables
In this section, we explain our variables. The descriptions are organized according to whether they
are independent, mediating or dependent variables in the SEM analyses.
52 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
2. Questionnaires from 103 US firms were returned as undeliverable, usually because the company had merged with
another.
5.2.1 Independent Variables
HR strategy: resource-based. There is no consensus on how to operationalize different types of HR
or business strategy empirically. Scholars have resorted to various indirect measures (Helfat
1997). Our approach is in this tradition of using proxy measures. We asked respondents to rate
the importance of six criteria in their company’s evaluation of the performance of the
headquarters HR function: level of direct and indirect labor costs, revenue or sales per em-
ployee, employee attitudes and morale, employee retention, co-operative relations with union
and union avoidance in unorganized facilities. Each indicator is rated on a 5-point scale ranging
from 0 (not important) to 4 (very important). We assumed that companies pursuing a resource-
based HR strategy would give priority to evaluation criteria incorporating the belief that
employees were a corporate asset. For Japan, we measured a resource-based HR strategy by the
average score on the indicators ‘employee attitudes and morale’ and ‘co-operative relations with
union,’ taking into consideration the prevalence of enterprise unions in large Japanese com-
panies. For the US, where only a minority of companies has unions (including in our sample),
we measured resource-based strategy by the score on ‘employee attitudes and morale.’
HR strategy: commodified. A different HR strategy is to minimize dependence on employees by
commodifying employment and rigorously minimizing labor costs. For both Japan and the US,
we measured the degree to which a company’s labor is commodified, by assessing the import-
ance of ‘level of direct and indirect labor costs’ and ‘revenue or sales per employee’ as criteria
for evaluating the performance of the headquarters HR function. We used the average score of
these two indicators on a 5-point scale ranging from 0 (not important) to 4 (very important). We
recognize the possibility that companies pursuing a resource-based HR strategy may achieve an
efficiency-wage effect by internally segmenting their workforce and treating some workers (‘core’
employees) as assets and other workers (temporary, part-time, outsourced) as commodities. In
this case, different HR strategies would be complements instead of substitutes (Ko 2003).
Corporate governance: stakeholder. To assess the strength of stakeholder corporate governance, we
measured the number of corporate directors who had prior managerial experience in the HR
area. While this is an indirect measure, it has the virtue of being objective and scalable. It is
based on the assumption that companies which value employees as stakeholders will prefer to
have individuals on their boards who are able to provide the employees’ perspective on major
business decisions. In Germany, this would occur via designated employee representatives. In
Japan and the US, we hypothesize that persons with an HR background are sensitive to
employee views and will present them to the board.
Corporate governance: shareholder. In Japan, the strength of shareholder corporate governance is
measured by adoption of the corporate officer system, which we code as 0 (J-model firms) or 1
(shareholder-oriented firms). In the US, we take a different approach and measure the strength
of a shareholder orientation by the use of stock options in the compensation of managers.
Companies that do not use stock options are coded as 0 and those that offer stock options to
at least some of their managers are coded as 1. We experimented with the stock option measure
for Japan but it had less explanatory power than the corporate officer system, perhaps because
options remain relatively uncommon in Japan.
5.2.2 Mediating Variables
Executive values: employee-oriented. We distinguish executive values that are relatively employee-
oriented vs. those that are relatively finance-oriented. As noted, we see these values as mediating
the relationship between governance and strategy on the one hand and corporate outcomes on
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 53
the other (Aguilera and Jackson 2003). We asked executives to rate on a 4-point scale (1=not
important, 4=most important) various issues and concerns in their job. We defined employee-
oriented values as the sum of scores on three items: safeguarding employees’ jobs, improving
employee morale and insuring that employees are treated fairly by line and senior managers.
Executive values: finance-oriented. Using the same rating scale, we measured finance-oriented values
as the sum of scores on the following three items: helping to raise shareholder dividends,
increasing the firm’s share price value and boosting the firm’s market share.
5.2.3 Dependent Variables
Internal labor markets. We hypothesize that internal labor markets (ILMs) are associated with
resource-based HR strategies and treating employees as stakeholders. We measured the strength
of ILMs in each company by examining its policies for filling vacancies for managerial and for
non-supervisory employees: whether they consider internal candidates only (coded as 2), give
first priority to internal candidates and recruit outside only when needed (coded as 1), consider
both internal and external candidates (coded as 0), or prefer recruiting external candidates
(coded as –2).3The ILM index is the sum of the measures for managerial positions and non-
supervisory positions respectively.
HR centralization. We assessed the degree to which operating authority over HR activities is
centralized, by asking respondents the percentage of responsibility assumed by different levels
(line managers, unit HR department, divisional HR department and headquarters HR depart-
ment) for five HR activities: employee participation plans, policies toward unions, decisions on
business unit headcount, job assignment of managers and performance evaluation of managers.
The mean percentage given for the headquarters HR department over the five activities was
used as the indicator of headquarters HR’s operating authority or centralization.
HR influence. Human resource’s influence is measured by the extent of HR’s involvement in two
key decisions: the selection and remuneration of senior executives and the allocation of payroll
budgets across divisions. We asked respondents what role they played in these decisions, whether
to provide information and/or to offer advice and/or take part in the final decision. For both
sets of strategic decisions, 1 point is given for providing information, 2 points for offering advice
and 3 points for taking part in the final decision, with the total ranging from 0 to 12 points.
HR relative power. We asked respondents to indicate the relative power to affect decisions at the
headquarters level of the following departments: finance, HR, marketing/sales, production,
planning and R&D, using a scale of 1–10. We then rank the departments according to the
rating (using mid-rank when there is a tie) and transform the rank to a relative rank by scaling
the highest-ranked department to have a relative rank of 1 and the lowest-ranked department
to have a relative rank of 0. Human resources power is measured by the relative rank of HR
among the rated headquarters departments (Perrow 1970).
6. Results
6.1 Longitudinal Data
Table 1 shows changes in the size of headquarters HR units in Japan and the US during the period
1996–2001. Most of the Japanese companies in our sample experienced flat or declining employment
54 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
3. For symmetry, we coded external candidates as –2, but it makes no difference to our results if coded as –1.
during this period, while cutting the number of headquarters HR staff. As a result, the number of
staff per employee fell from 1 for every 106 employees in 1996 to 1 for every 129 employees in
2001. Conversely, most of the US companies experienced employment growth and added modestly
to their headquarters HR staff. However, the rate of staff growth was much lower than the rate of
employment growth, so that the ratio of staff per employee actually declined from 1996 to 2001.
Hence, one can interpret Table 1 as supporting convergence (the percentage change in staff) or as
inconsistent with convergence (because the Japan–US gap in staff-per-employee widened during the
period).
The HR function’s relative share of total headquarters staff also is of interest, so we asked
respondents to tell us whether this share had decreased, stayed the same or increased during the
period 1996–2001, as shown in Table 2. Again, evidence relevant to convergence is mixed. On the
one hand, US respondents were more likely to say that HR’s share of headquarters staff had grown
larger. On the other hand, a large proportion of US respondents reported that HR’s relative share
of headquarters staff was declining; this despite the fact that employment at many of these
companies was increasing.
We gave respondents a list of seven headquarters functions and asked whether each had gained or
lost power to influence strategic decisions during the previous five years. In Table 3, we present the
top three ‘gainers’ among the headquarters functions. In Japan, despite cuts in HR staff, HR ranked
second (40%)—behind the planning department—in the percentage of respondents saying that it
had gained power during the previous five years. Finance was ranked third. In the US, the HR
respondents were overwhelming in their claim that HR gained power (77%), followed by finance
and planning. One can interpret these findings as supportive of convergence, as US HR executives
judge themselves as gaining power more rapidly than their Japanese counterparts.
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 55
Table 1. Change in Size of Headquarters HR Units.
1996–2001 1996 2001
% change in number of No. of headquarters HR
headquarters HR staff staff per employee
Japan –22 1/106 1/129
US 4 1/144 1/185
Table 2. Change in HR’s Share of Headquarters Employment.
1996–2001 (% of companies)
Smaller Same Larger
Japan 47 47 5
US 40 35 24
We asked respondents to evaluate changes over time in line-manager involvement in personnel
decisions: performance evaluation and job assignment of managers, control over employee partici-
pation, decisions on headcount and policies toward unions. We view line-manager involvement as a
proxy for a decentralized, market-oriented approach to employment. As shown in Table 4, the
strongest tendency in both countries is to say that there was no change during the 1996–2001
period. However, US executives were more likely than Japanese executives to report an increase in
line involvement in every category of decision-making, with the exception of union policies. The
finding suggests that, although there is decentralization in both countries, the US is decentralizing
more rapidly. Given the likely fact that US firms were more decentralized to begin with, the finding
suggests a widening of the central-tendency gap between Japanese and US companies, even though
at the margin they are moving in the same direction.
Finally, we replicated items from a 1993 survey of Japanese directors which asked respondents to
judge what was important to them in their jobs on a scale of 1–4, with 4 the most important. As
shown in Table 5, our data are limited to HR executives, whereas the earlier data are for board
56 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
Table 3. Change in Strategic Influence.
1996–2001
Function gaining power % of respondents
Japan Planning 54
HR 40
Finance 37
US HR 77
Finance 50
Planning 35
Table 4. Change in Line Involvement.
1996–2001 (% of companies)
Increased Same Decreased
Japan US Japan US Japan US
Performance evaluation of managers 39 53 57 43 4 4
Job assignment of managers 29 40 63 52 8 7
Introduce or modify participation plans 23 44 66 52 11 4
Decisions on business unit headcount 21 46 72 46 8 8
Develop policies toward unions 18 15 75 76 6 9
members regardless of function. One might expect the latter group to be more ‘pro-shareholder’
than the HR executives. Hence, the increase in importance accorded to share price (line 2) and the
decrease in importance attached to safeguarding employee jobs (line 5) may be understated.
Nevertheless, what is striking about Table 5 is the small magnitude of the absolute and relative
changes in Japanese executive values between 1993 and 2001. Safeguarding employee jobs remains
far and away of great importance compared with share price (or dividends or market share, which
have declined in importance) and also compared with the values of US executives, for whom
dividends, share price and market share are more important than safeguarding jobs. In short,
attitudinal convergence might be occurring, but the evidence for it in Table 5 is weak.
6.2 Cross-sectional Data
Next, we compare national variable means as an indirect measure of convergence. If there had been
substantial convergence, variable means would not be significantly different between Japan and the
US. Table 6 reports the means, standard deviations and correlation coefficients of the variables.
Summary statistics for Japan are shown below the diagonal and those for the US, above the
diagonal.
We ran two-tailed t-tests with unequal variances, to see whether there are significant differences
between the Japan and US sample means. The means of the variables are significantly different at
the 0.0001 level with the exception of the HR power variable, which is significant at the 0.15 level.
Note that the variable means are as one would predict: Japanese firms score more highly on
resource-based HR strategies, employee-oriented executive values, the ILM index and on HR
centralization, HR influence and HR relative power; the US firms score more highly on finance-
oriented values. Indeed, the data show that the most powerful department in the US was finance,
with HR ranked fifth, whereas in Japan, the top departments were planning and marketing, with HR
and finance tied for third. Thus, the national averages reflect the dominance of the resource-based,
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 57
Table 5. Executive Values in Japan and the US.
1993 2001 2001
Japanese Japanese HR US HR
Directors Executives Executives
Raising dividends 2.6 2.2 2.6
Share price 2.0 2.3 3.3
Market share 2.9 2.2 2.9
Diversify & expand into new markets 2.9 2.5 2.4
Safeguard employees’ jobs 3.3 3.2 2.1
Increase number of management positions 1.3 1.2 1.2
Increase my department’s budget 1.5 1.4 1.3
Coordinate with other departments 2.4 2.8 3.2
Make contribution to society 2.6 2.5 2.4
Note: ‘What is important to you in your job?’ 1=not important, 4=most important.
1993 data courtesy of Suzuki Fujikazu, RENGO Research Institute for Advancement of Living Standards (RIALS), Tokyo.
58 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
Table 6. Variable Means, Standard Deviations, and Correlations.
12 3 4 56 789101112 U.S. U.S.
Mean s.d.
1HR strategy: –0.05 –0.01 – 0.25* 0.26** 0.07 0.08 –0.16†–0.01 0.10 –0.07 3.09 0.89
resource-based
2HR strategy: 0.22** 0.13 – –0.02 –0.00 0.15 0.17†0.02 0.13 0.08 0.04 2.11 0.90
commodified
3Corporate governance: 0.11 –0.04 – 0.08 0.05 0.07 0.17†–0.11 –0.10 –0.13 0.11 0.49 0.81
stakeholder
4Corporate governance: –0.07 0.08 –0.09 – – – – – – – – – –
shareholder (Japan)
5Corporate governance: –0.26*** 0.04 0.07 0.26*** 0.21* 0.17†–0.08 –0.05 0.07 0.08 0.05 0.98 0.14
shareholder (U.S.)
6Executive values: 0.24** 0.10 0.04 –0.02 –0.08 0.15 0.19†–0.08 –0.02 –0.10 –0.14 8.80 1.50
employee-oriented
7Executive values: 0.09 0.15* –0.01 0.03 0.11 0.04 –0.09 –0.08 0.15 0.05 –0.03 8.76 2.23
finance-oriented
8Internal labor markets 0.16* 0.08 –0.05 –0.01 –0.07 0.07 –0.03 –0.13 0.17†0.08 0.02 1.02 0.89
9HR centralization –0.02 –0.05 0.08 –0.20** –0.15* –0.15* 0.13†0.16* 0.01 0.05 –0.10 28.6 21.6
10 HR influence 0.08 0.09 –0.08 0.01 –0.03 0.02 0.12†–0.5 0.03 0.21* 0.13 5.17 2.21
11 HR relative power 0.09 –0.10 0.19** 0.05 0.02 –0.06 0.08 0.06 0.06 –0.07 0.19* 0.35 0.30
12 Reporting to CEO (U.S.) – – – – – – – – – – – 0.66 0.48
Japan Mean 3.52 3.05 1.67 0.29 0.17 9.82 6.72 2.40 48.8 6.45 0.41 –
Japan s.d. 0.61 0.79 1.71 0.45 0.38 1.37 2.06 1.10 18.3 2.22 0.27 –
Notes: Statistics for Japan shown below the diagonal and those for the US above the diagonal.
n(Japan)=181, n(US)=105.
† p<0.10; * p<0.05; ** p<0.01; *** p<0.001.
stakeholder and strong-HR model in Japan and of the shareholder-oriented and weak-HR model in
the US. These findings are consistent with non-convergence of corporate systems, although it is
possible that variable means were more disparate in the past than now.
The exception here is that Japanese companies score more highly on commodified HR strategies
(that is, they are more concerned with reducing labor costs) than are US firms. One interpretation
of this finding is that Japanese companies are trying to preser ve a high-wage, stable-employment
core by internally segmenting their workforces and expanding the number of contingent and
overseas employees. Hence, they are pursuing both resource-based and commodified labor strategies.
Another interpretation is cyclical: the different emphasis on cost-cutting is due to the effect of the
Heisei recession in Japan and to the fact that the survey was conducted at the tail end of the US
stock-market bubble, when costs were of secondary concern in many US companies.
6.3 Structural Equation Modeling Analyses
Structural equation modeling is an extension of the general linear model of which multiple regres-
sion is a part. SEM serves purposes similar to multiple regression, but is a more powerful alternative
in particular situations. Advantages of SEM compared with multiple regression include the testing
of overall models rather than individual coefficients, the ability to test models with multiple dependent
variables and the ability to model mediating variables. While some of these features can be adapted
to a multiple regression framework, SEM is able to handle all of them simultaneously. Given the inter-
dependencies of various aspects of corporate organization, SEM is particularly suited to our analysis.
For both the Japan and US datasets, the parameter estimates obtained by using maximum-
likelihood (ML) estimation would be biased owing to excessive multivariate kurtosis.4To deal with
this problem, we fit the models using robust ML methods and corrected for kurtosis using the
Satorra–Bentler (1988) method. For each country, the sample size is large enough to allow for
reliable parameter estimates and significance tests for models of the size depicted in Figure 1.
In what follows, we first present overall fit information for the concurrent existence of two models
in contemporary Japan: the postwar model and the US-style model. Next, we present our results on
specific path relationships. Then we follow the same procedure for the US.
6.3.1 Japan
Fitting the hypothesized two-model structure to the data yields no clear assessment of fit (Figure 2).
The chi-square test for exact fit is significant (chi-square (30 df)=46.07, p=0.031), making it
necessary to rely on the customary indices for fit. Most of them indicate close fit: GFI=0.954;
SRMR=0.066; and RMSEA=0.055, with 90% confidence interval of (0.017, 0.084). Only the
CFI=0.622 is below the customary cutoff score (Hu and Bentler 1995).
We decided to modify the model by removing the shareholder corporate governance variable
(corporate officer system), which is not significantly related to any other variable. It may well be that
recent changes in Japanese corporate governance such as the corporate officer system are largely
symbolic—intended to signal to foreign investors that management is sensitive to shareholders’
interests—and as yet have not affected fundamental corporate processes. There is evidence that US
managers are prone to making symbolic changes in corporate governance to satisfy outside con-
stituents, so it would not be surprising to observe the same behavior in Japan (Powell and DiMaggio
1983; Westphal and Zajac 1998).
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 59
4. Normalized Mardia’s kappa=7.62 for Japan and 12.04 for the US.
Removing this variable improves fit to an acceptable level. The chi-square test for exact fit is non-
significant (chi-square (25df)=34.42, p=0.099), indicating that there is no difference between the
data matrix implied by the model and the sample data. All fit indices mirror this improved fit by
increasing even further: GFI=0.962; SRMR=0.063; and RMSEA=0.046 with a 90% confidence
interval of (0.000, 0.080). The CFI is still below the customary cutoff (CFI=0.747) but, as all the
other fit indices agree in the assessment that this model fits the data well, and—more importantly—
because of the non-significance of the chi-square test, we feel it is warranted to accept this model as
a good representation of the data.
As regards path relationships, the modified Japan model provides weak support for the existence
of a positive relationship between resource-based HR strategy and stakeholder corporate govern-
ance. We find a significant, positive association between stakeholder corporate governance and the
60 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
Figure 2. Parameters for Theoretical and Modified Models for Japan.
Note: Parameters in parentheses are for the theoretical model and those in bold are for the modified model.
Variables and paths with dotted lines are removed in the modified model.
power of the headquarters HR function. Stakeholder governance is positively associated with HR
executives having employee-oriented values, although the path is not significant. Executive values
also are significantly associated with a company’s HR strategy, whether the pairing is between a
resource-based strategy and employee-oriented values or between a commodified-labor strategy and
finance-oriented values. However, our hypothesis of a positive association between stakeholder corporate
governance and HR influence is not supported: the sign is negative.
As for the paths between executive values and organizational outcomes, the results are mixed. Our
hypothesis—that ILMs are positively related to employee-oriented values but negatively related to
finance-oriented values—came out with the correct signs but neither path was significant. We
thought that HR centralization would be positively related to employee-oriented values, but the
observed relationship is negative and significant. And rather than there being a negative association
between centralization and finance-oriented values, we found the observed relationship to be positive
and significant. Finally, we found, as predicted, that centralization is positively associated with ILMs.
Stepping back from the myriad paths, the larger point is this: the postwar J-model continues to
exist, and there is also mild evidence consistent with the emergence of a US-style model.
6.3.2 United States
For the US analyses, we added an additional variable not contained in the Japanese dataset:
reporting to CEO. The variable distinguishes between HR executives who report to the CEO
(coded as 1) and those who do not (coded as 0). Fitting the hypothesized model to the data yielded
(see Figure 3) a non-significant chi-square (chi-square (38 df)=47.83, p=0.13), which means there
is no difference between the values predicted by this model and the sample data. The conventional
fit indices mirror this result: GFI=0.922; SRMR=0.082; and RMSEA=0.050 with a 90% confidence
interval of (0.000, 0.089). The only outlier, again, is the CFI=0.638 but, given the result of the chi-
square test, the CFI does not affect our judgment that the model is a good representation of the
data.
Figure 3 shows path estimates for the US. Stakeholder governance is positively but not signifi-
cantly associated with employee-oriented values, although there is a strong positive relationship
between a resource-based HR strategy and employee-oriented values. As for finance-oriented values,
they are strongly related to shareholder corporate governance, as predicted. Also, there is a positive
relationship between a commodified-labor HR strategy and finance-oriented executive values, as
predicted.
However, we found no relationship between a company’s HR strategy and its corporate
governance system, whether stakeholder or shareholder. Also, neither of the paths from stakeholder
governance to HR influence and to HR relative power is positive; they are negative and significant.
Nor do we find support for a negative relationship between shareholder governance and HR
influence and HR relative power: the paths are positive and significant. We shall return to this point.
As for the paths between executive values and organizational outcomes, the results are mixed. As
predicted, ILMs are positively and significantly related to employee-oriented values and negatively
related to finance-oriented values. However, neither of the paths between values and HR central-
ization is significant. Reporting to the CEO is positively associated with HR influence and with HR
relative power, as predicted and the latter path is significant.
Again, stepping back to see the larger picture, the strong fit of our SEM model is consistent with
the concurrent existence of a traditional US-model and of an emerging J-model in US companies,
despite the insignificance of some particular paths.
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 61
7. Discussion
The SEM results provide evidence both for supporters and skeptics of the notion that corporate
systems in Japan and the US are converging. There are two types of evidence relevant to conver-
gence: the similarity of pairwise relationships in our SEM analyses and of the groups of relationships
or overall models. Regarding the former, note that in both Japan and the US, executive values are
aligned with HR strategy, which could either be due to companies choosing HR executives with the
‘right’ set of values or to HR executives being sensitive to organizational decisions about how to
achieve competitiveness. Secondly, in both countries HR strategy and corporate governance are not
significantly related. It appears to be the case that corporate governance can coexist with different
types of HR strategies, such that there is not a well-defined relationship between them. Thirdly, in
both countries the HR executive’s values are related to the strength of internal labor markets.
Human resources executives with employee-oriented (finance-oriented) values tend to be found in
companies with relatively strong (weak) internal labor markets.
62 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
Figure 3. Parameters for Theoretical Model for the US.
Earlier, we posited that there were different models for organizing the corporation: the J-model
of the resource-based/stakeholder corporation and the US model of the commodified-
labor/shareholder corporation. Do these models exist, and do they have similar effects in both
countries?
In Japan, there is evidence of the continued existence of the postwar resource-based/stakeholder
model, which is associated with HR relative power and with internal labor markets. However,
evidence for the existence in Japan of a US-style shareholder model is weaker. Recall that we
removed shareholder governance from our model because it was not significantly related to other
variables. There is, however, a set of paths running from a commodified-labor strategy to finance-
oriented values to centralization. These paths may constitute an emerging alternative to the postwar
Japanese model. Alternatively, these paths may represent the emergence of a hybrid system in Japan,
in which more market-oriented employment and shareholder sensitivity are added on to the postwar
system. Indeed, as shown in Table 6, commodity-based and resource-based strategies are positively
and significantly related. This may indicate an intensified core-periphery approach inside large
Japanese corporations, in which resource-based and commodified HR strategies are pursued as
complements rather than substitutes.
Also, note that in Japan, corporate governance is not as strongly associated with values and
outcomes as in the US. For Japan, the association is slightly stronger between strategy and outcomes
than between governance and outcomes. Admittedly, this is a thin empirical reed. But it suggests
that perhaps too much scholarly and journalistic attention is being devoted these days to changes in
Japanese corporate governance (the question of how firms make decisions) and not enough atten-
tion is being paid to variations in HR and business strategy (the actual content of those decisions).
Next, we turn to the US, for which the data support the existence of two different corporate
models: the traditional US model and something resembling the postwar Japanese model, with its
links between resource-based strategy, pro-employee values and employment security. However, our
hypothesis that the latter model is positively associated with HR power (as it is in Japan) was wrong.
Conversely, our hypothesis that the commodified-labor/shareholder model is negatively associated
with HR power and influence also was wrong. Rather, the association is positive, a finding that lends
credence to the claim—which initially we were skeptical about—that the US is developing a
prestigious ‘business partner’ role for the HR executive, despite shareholder sovereignty and com-
modified employment. Human resources executives who adapt to this role by assuming finance-
oriented values are able to achieve power and influence within their organizations.
It is unclear why the stakeholder approach is not associated with HR power in the US. One answer
may lie in our data showing that reporting to the CEO, ceteris paribus, has a significant positive
effect on HR’s strategic influence and relative power. A recent study finds that CEO reporting is
most prevalent in companies that have restructured by downsizing their middle-management ranks
(Rajan and Wulf 2003). Human resources executives might incline themselves to a powerful
downsizing-oriented CEO rather than to any stakeholder elements on the board, where persons
with an HR background are less numerous than in Japan. Future research on both countries should
pay closer attention to the dynamics of board decision-making.
8. Conclusions
This paper has analyzed corporate organization in Japan and the US to determine whether there is
evidence of convergence of those systems. We presented longitudinal and cross-sectional data as well
as SEM analyses of corporate organization. The results provide evidence for both sides in the convergence
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 63
debate: those who see an erosion of national systems and those who contend that ‘varieties of
capitalism’ remain distinctive.
In support of convergence were the following: (a) the longitudinal data show a shrinkage of
headquarters HR units in Japan and the growth of HR power in the US; (b) the cross-sectional data
confirm the existence of commodified-labor HR strategies in Japan and of resource-based HR
strategies in the US; and (c) the SEM analyses are consistent with the existence of a J-model in the
US and also show cross-national similarities in the relationships between variables, such as the
alignment of HR executive values with strategy and employment practices.
Ideas about corporate organization are flowing across national borders, and the flows go in both
directions. Japan is pursuing more commodified-labor strategies, while US companies are pursuing
resource-based HR strategies and stakeholder governance. Thus, what we observe is not simply that
Japan is becoming more like the US, but that the US is changing too. This is bidirectional rather
than US-centric convergence, and we urge that more attention be paid to flows in both directions.
In support of non-convergence were the following: (a) the longitudinal data show the persistence
of pro-employee executive values in Japan and a widening of HR centralization and HR staffing
ratios between Japan and the US; (b) the cross-sectional data demonstrate statistically significant
differences between Japan and the US in most measures of corporate organization; and (c) the SEM
results show the structural weakness of the shareholder model in Japan. Also, although there is
evidence of a resource-based approach in the US, it is not the route to executive HR power, as it is
in Japan.
One barrier to convergence is the difficulty of creating governance forms that can sustain new
approaches. In the US, it is curious that stakeholder governance and a resource-based HR strategy
do not reinforce each other and fail to bolster HR power and influence. Others have made the point
that an HR strategy based on high human-capital investments is best supported by stakeholder-
oriented corporate governance (Roberts and Van den Steen 2000). As yet, this orientation is rather
scarce in the US. Although having individuals on corporate boards with HR backgrounds inclines a
company in the stakeholder direction, a critical mass does not yet exist on most US boards. The
number of such directors is less than one-third of Japanese levels. Recently, the AFL-CIO and
shareholder activists such as CalPERS have pressured the Securities Exchange Commission (SEC)
to permit shareholders to nominate board directors (Plitch 2003). If successful, this might push
boards in a stakeholder direction. Ironically, it would occur at precisely the same time that Japanese
law is changing to facilitate a shareholder orientation.
A barrier to convergence on the Japanese side is the fact that, as yet, most Japanese companies
have not ‘financialized’ themselves. The finance function does not dominate operational or strategic
decisions. The majority of Japanese companies have not adopted the corporate officer system or the
new ‘American’ system of outside directors, so they continue to have the HR function represented
on their boards. And even some corporate-officer companies still have an HR director or other
board members with HR backgrounds. Contrast this to the US, where only six of the Fortune 1000
US companies have their own HR executive on the board of directors, whereas 92 companies have
their CFO on the board.5
64 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
5. Unpublished Korn/Ferry data as of 2002, courtesy of Caroline Nahas. Although much is made of the fact that
independent directors are becoming prevalent in Japan, only one-third of Japanese companies have these directors and
more than half of those companies have but one independent director (Nikkei Weekly 2004). Again, this suggests the
often symbolic nature of changes in corporate governance.
It is impossible to make sense of these impediments to convergence without taking into
consideration issues of power, both within and outside the corporation. There is a tradition in
organizational research (Pfeffer 1981) that views differences in corporate organization as being due
not only to efficiency considerations but also to contests over the distribution of resources. Functional
units inside the firm—such as finance, marketing and HR—vie for power, not only for themselves,
but as representatives of different groups, such as shareholders, customers and employees (Freeland
2001).
Paying greater attention to power is important for understanding the varieties of capitalism
debate. National models are due, in part, to different allocations of power between stakeholders,
and these distributions generate distinctive organizational arrangements. Seeing things from a
power perspective sensitizes us to the possibility that debates over the relative efficacy of different
corporate models are, at a latent level, really disputes over distributional outcomes. We urge other
researchers to give more consideration to how corporate organization is affected by power contests
within corporations and the societies in which they are embedded. We do not view our research as
the final word on this subject but instead see it as a first step toward operationalizing and measuring
issues that are being hotly debated—and need more empirical analysis—in both Japan and the US.
References
Abe, Masahiro. 2002. ‘Corporate Governance Structure and Employment Adjustment in Japan: An Empirical Analysis
Using Corporate Financial Data’. Industrial Relations 41: 683–702.
Agle, Bradley, Ronald K. Mitchell, and Jeffrey A. Sonnenfeld. 1999. ‘Who Matters to CEOs? An Investigation of
Stakeholder Attributes and Salience, Corporate Performance, and CEO Values’. Academy of Management Journal 42:
507–525.
Aguilera, Ruth V. and Gregory Jackson. 2003. ‘The Cross-National Diversity of Corporate Governance’. Academy of
Management Review 28: 447–85.
Ahmadjian, Christina L. 2002. ‘Changing Japanese Corporate Governance’. In Japan’s Managed Globalization:
Adapting to the 21st Century, ed. Ulrike Schaede and William A. Grimes. Armonk: M.E. Sharpe.
Aoki, Masahiko. 1988. Information, Incentives and Bargaining in the Japanese Economy. Cambridge: Cambridge
University Press.
Applebaum, Eileen, Thomas Bailey, Peter Berg, and Arne Kalleberg. 2000. Manufacturing Advantage: Why High
Performance Work Systems Pay Off. Ithaca: Cornell University Press.
Araki, Takashi. 2000. ‘A Comparative Analysis: Corporate Governance and Labor and Employment Relations in Japan’.
Comparative Labor Law & Policy Journal 22: 67–96.
Barney, Jay. 1991. ‘Firm Resources and Sustained Competitive Advantage’. Journal of Management 17: 99–120.
Bebchuk, Lucian A., Jesse Fried, and David I. Walker. 2002. ‘Managerial Power and Rent Extraction the Design of
Executive Compensation’. University of Chicago Law Review 69:751–846.
Bird, Allan and Schon Beechler. 1995. ‘Links Between Business Strategy and Human Resource Management’. Journal
of International Business Studies 26: 23–48.
Blair, Margaret M. 1995. Ownership and Control: Rethinking Corporate Governance for the Twenty-First Century.
Washington, D.C.: The Brookings Institution.
Cappelli, Peter. 1999. The New Deal at Work: Managing the Market-driven Workforce. Boston: Harvard Business School Press.
Cassidy, John. 2002. ‘The Greed Cycle’. The New Yorker (23 September): 64–77.
Chandler, Alfred D. 1962. Strategy and Structure: Chapters in the History of the American Industrial Enterprise.
Cambridge: MIT Press.
Cole, Robert E. 1999. Managing Quality Fads: How American Business Learned to Play the Quality Game. New York:
Oxford University Press.
Cronin, James E. 2000. ‘Convergence by Conviction: Politics and Economics in the Emergence of the Anglo-American
“Model”’. Journal of Social History 33: 781–804.
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 65
Csoka, Louis S. 1995. Rethinking Human Resources. Report number 1124-95-RR. New York: The Conference Board.
Dobbin, Frank and John R. Sutton. 1998. ‘The Strength of a Weak State: The Rights Revolution and the Rise of
Human Resources Management Divisions’. American Journal of Sociology 104: 441–476.
Dore, Ronald. 2000. Stock Market Capitalism: Welfare Capitalism: Japan and German versus the Anglo-Saxons. Oxford:
Oxford University Press.
Fligstein, Neil. 1987. ‘The Intraorganizational Power Struggle: Rise of Finance Personnel to Top Leadership in Large
Corporations’. American Sociological Review 52:44–58.
Freeland, Robert. 2001. The Struggle for Control of the Modern Corporation: Organizational Change at General Motors,
1924–1970. Cambridge: Cambridge University Press.
Fruin, W. Mark. 1994. The Japanese Enterprise System. Oxford: Oxford University Press.
Gordon, Andrew D. 1998. The Wages of Affluence: Labor and Management in Postwar Japan. Cambridge: Harvard
University Press.
Groshen, Erica L. 1991. ‘Five Reasons Why Wages Vary Among Employers’. Industrial Relations 30: 350–371.
Guillen, Mauro F. 2001. The Limits of Convergence. Princeton: Princeton University Press.
Hall, Peter A. and David W. Soskice. 2001. Varieties of Capitalism: Institutional Foundations of Comparative
Advantage. New York: Oxford University Press.
Helfat, Constance. 1997. ‘Know-how and Asset Complementarity and Dynamic Capability Accumulation: The Case
of R&D’. Strategic Management Journal 18: 339–60.
Holmes, Stanley and Wendy Zellner. 2004. ‘The Costco Way’. Business Week (12 April): 76.
Hu, Li-Tze and Peter M. Bentler. 1995. ‘Evaluating Model Fit’. In Structural Equation Modeling: Issues and
Applications, ed. Rick H Hoyle. Thousand Oaks, CA: Sage: 76–99.
Imai, Kenichi and Hiroyuki Itami. 1984. ‘Interpenetration of Organization and Market’. International Journal of
Industrial Organization 2: 285–310.
Inagami, Takeshi and the Research Institute for the Advancement of Living Standards. 2000. Gendai Nihon no
Koporeto Gabanansu (Corporate Governance in Contemporary Japan). Tokyo: ToyoKeizai.
Inohara, Hideo. 1990. The Japanese Personnel Department: Structure and Functions. Tokyo: Sophia University.
Itami, Hiroyuki. 1987. Mobilizing Invisible Assets. Cambridge: Harvard University Press.
Itoh, Hideshi. 1994. ‘Japanese Performance Management from the Viewpoint of Incentive Theory’. In The Japanese Firm:
The Sources of Competitive Strength, ed. Masahiko Aoki and Ronald Dore. Oxford: Oxford University Press: 232–264.
Jacoby, Sanford M. 2004. The Embedded Corporation: Corporate Governance and Employment Relations in Japan and the
United States. Princeton: Princeton University Press.
Janger, Allen R. 1977. The Personnel Function: Changing Objectives and Organization. New York: Conference Board.
Kagono, Tadao, Ikujiro Nonaka, Kiyonori Sakakibara, Shiori Sakamoto, and Johnny K. Johansson. 1985. Strategic
versus Evolutionary Management: A US-Japan Comparison of Strategy and Organization. New York: Elsevier.
Kato, Takao. 2001. ‘The End of “Lifetime Employment” in Japan? Evidence from National Surveys and Field Research’.
Working Paper, Colgate University, Hamilton, NY.
Katz, Harry C. and Owen Darbishire. 2000. Converging Divergences: Worldwide Changes in Employment Systems.
Ithaca: ILR Press.
Katz, Richard. 2003. Japanese Phoenix: The Long Road to Economic Revival. Armonk: M.E. Sharpe.
Khurana, Rakesh. 2002. Searching for a Corporate Savior: The Irrational Quest for Charismatic CEOs. Princeton:
Princeton University Press.
Kletzer, Lori G. 1998. ‘Job Displacement’. Journal of Economic Perspectives 12: 115–136.
Ko, Jyh-Jer Roger. 2003. ‘Contingent and Internal Employment Systems: Substitutes or Complements?’ Journal of Labor
Research 24: 473–90.
Kono, Toyohiro. 1999. ‘A Strong Head Office Makes a Strong Company’. Long Range Planning 32: 225–236.
Kramer, Robert J. 1999. Organizing for Global Competitiveness: The Corporate Headquarters Design. Research Report
1233-99-RR. New York: The Conference Board.
Levine, David I., Dale Belman, Gary Charness, Erica Groshen, and K. C. O’Shaughnessy. 2003. Changes in Careers
and Wage Structures at Large American Employers. Kalamazoo: Upjohn Institute.
66 Sanford M. JACOBY, Emily M. NASON and SAGUCHI Kasuro
Lieberman, Marvin B. and Rajeev Dhawan. 2001. ‘Assessing the Resource Base of Japanese and US Auto Producers’.
Working Paper, Anderson School at UCLA, Los Angeles.
McFarland, Dalton E. 1962. Cooperation and Conflict in Personnel Administration. New York: American Foundation for
Management Research.
Mowery, David C. and Richard R. Nelson. 1999. Sources of Industrial Leadership: Studies of Seven Industries.
Cambridge: Cambridge University Press.
Nikkei Weekly.2004. ‘More Firms Tap Outside Directors’ (13 September): 3.
Perrow, Charles. 1970. ‘Departmental Power and Perspectives in Industrial Firms’. In Power in Organizations, ed.
M. N. Zald. Nashville: Vanderbilt University Press.
Pfeffer, Jeffrey. 1981. Power in Organizations. Marshfield: Pittman Publishing.
———.1998. The Human Equation: Building Profits by Putting People First. Boston: Harvard Business School Press.
Plitch, Phyllis. 2003. ‘Investors Seek Proxy Election Reform’. Los Angeles Times (17 February): C3.
Powell, Walter W. and Paul J. DiMaggio. 1983. ‘The Iron Cage Revisited: Institutional Isomorphism and Collective
Rationality in Organizational Fields’. American Sociological Review 48: 147–60.
Price, Karl F. and James W. Walker, eds. 1999. The New HR: Strategic Positioning of the Human Resource Function.
New York: Human Resource Planning Society.
Rajan, Raghuram and Julie Wulf. 2003. ‘The Flattening Firm: Evidence from Panel Data on the Changing Nature of
Corporate Hierarchies’. Working Paper 9633. Cambridge, MA: National Bureau of Economic Research.
Reichheld, Frederick F. 1996. The Loyalty Effect. Boston: Harvard Business School Press.
Ritzer, George and Trice Harrison. 1969. An Occupation in Conflict: A Study of the Personnel Manager. Ithaca:
New York State School of Industrial & Labor Relations Press.
Roberts, John and Eric Van den Steen. 2000. ‘Shareholder Interests, Human Capital Investment, and Corporate
Governance’. Working Paper, Stanford Graduate School of Business.
Rush, Harold M. F. 1969. Behavioral Science: Concepts and Applications. New York: National Industrial Conference
Board.
Satorra, Albert and Peter M. Bentler. 1988. ‘Scaling Corrections for Chi-square Statistics in Covariance Structure
Analysis’. In American Statistical Association, Proceedings of the Business and Economics Section. Alexandria, VA:
American Statistical Association: 308–313.
Tachibanaki, Toshiaki and Atsuhiro Taki. 2000. Capital and Labour in Japan: The Functions of Two Factor Markets.
London: Routledge.
Troy, Mike. 2003. ‘Seiyu Adopts Wal-Mart’s Characteristics’. DSN Retailing Today (22 September): 5, 34.
Ulrich, David. 1997. Human Resource Champions: The Next Agenda for Adding Value and Delivering Results. Boston:
Harvard Business School Press.
Westphal, James and Edward Zajac. 1998. ‘The Symbolic Management of Stockholders: Corporate Governance Reforms
and Shareholder Reactions’. Administrative Science Quarterly 43: 172–53.
Wright, Patrick M., Benjamin Dunford, and Scott Snell. 2001. ‘Human Resources and the Resource-based View of
the Firm’. Journal of Management 27: 701–21.
Corporate Organization in Japan and the United States: Is There Evidence of Convergence? 67
