TSA04.rtf

中文摘要

制度鑲嵌與企業聯盟網絡的形成：一個慣時性的研究

近年來市場與組織社會學家對於企業組織的競爭與合作行為有了大量而深入的研究，其中一個研究主

題是企業的聯盟網絡的形成。本研究接續這個研究主題，並進一步探討，企業的競爭與合作如何被市場網絡結

構、以及更大的制度環境所影響；尤其是後者，雖然一直被學者倡議，卻仍是一個待被回答的研究主題。而

1990 年代以降的台灣社會，在政治與經濟自由化之後經歷劇烈的制度環境變革，並衍生出許多新的市場機會，

引發了一連串的版圖重建與緩衝市場衝擊的企業聯盟熱潮，這個商業活動的趨勢成為本研究測試的一個極佳個

案。本文的基本假設是，在市場重組與制度環境的變動中，企業既有的制度鑲嵌具有正當性背書的功效，也成

為企業在茫茫市場中選擇聯盟伙伴的重要判準。藉著整合制度論、地位模型

(Podolny 1993) 以及組織範疇認同

的理論

(Zuckerman 1999)，本研究導引出討論制度鑲嵌對聯盟網絡影響的幾個研究假設，並以 1990 至 2003 年

的

211 個台灣上市公司的股權聯盟活動進行統計分析。研究結果發現制度鑲嵌對於企業聯盟行為有正面的影響

效果，雖然並不是單純的線性關係：過度的制度鑲嵌反而帶來負面效果，這與

Uzzi 的市場鑲嵌研究有異曲同

工之妙。而研究結果也發現，當企業所身處的產業在聯盟活動上由於過度分散而被視為不佳的聯盟伙伴時，制

度鑲嵌卻有緩和這種不利的效果。此外，在市場隨著經濟環境自由化而愈形階層化的同時，制度鑲嵌也與市場

階層化的力量互相強化而影響企業聯盟的形成。這些發現不止推進了社會學家對於企業聯盟行為的進一步瞭

解，同時也動態的認識到制度環境與市場網絡對於企業競爭行為的交互影響與限制。

INTRODUCTION

In the past two decades, one vigorous practice within sociology is the study on economic

markets in which scholars have labored assiduously to apply models of social structure to analyze
corporate behavior. One fundamental assumption underlying such a body of works is that
corporate activities are determined by the broader social contexts and institutional environments
in which corporate organizations are embedded (Granovetter 1985). Important pioneering works
echoing this theme include Harrison White’s analysis of production markets (1981; 2002), Ronald

Burt’s (1992) “structural hole” theory and Wayne Baker’s (1984) analysis of price volatility on the
trading floor. Significant follow -ups that add a flavored twist on the embeddedness theme also
include Joel Podolny’s “status-based” model, and Ezra Zuckerman’s (1998) attempts to synthesize
it with the Durkeimian idea of categorical imperative in social life. Aiming at rejecting the pure
economistic line of reasoning, or, a so called “under-socialized” perspective, sociologists have
achieved better insight into the operation of modern corporations and the social-structural
mechanisms sustaining them.

Within copious amounts of sociological studies of markets, one recurrent topic that appeals

to scholars are the dynamic activities of intercorporate alliances. A strategic alliance is usually
defined as a voluntarily initiated contractual asset pooling or resource exchange agreements
between firms (Gulati 1999; Stuart 1998). While most of the work seeks to understand alliance
behavior with attribute-based (i.e., “resource based”) explanations – and is therefore devoted to
many of the efforts in identifying how interorganizational coalitions are predisposed by firms’
characteristics, such as their size or financial condition (e.g., Wernerfelt, 1984; Barney, 1991;
Mahoney and Pandian, 1992) – sociologists tend to question this very reasoning and ask how the
availability of and access to alliance opportunities are conditioned by firm’s position in its external
environment. For example, Gulati (1995a) has brought the embeddedness perspective into alliance
studies and argued that social networks play a significant role in facilitating new alliances by
providing valuable information to firms about the specific capabilities and reliability of potential
partners. In a later work (1999) he also demonstrated that social ties and cohesive groups in which

firms connected are important resource channels in gaining entry into alliances. Similarly, when
considering the fact that business networks were mostly knotted repeatedly, Gulati and Gargiulo

(1999) argued that corporate alliances is an iterative process in which new partnerships
dynamically modify the previous ones, which then, again, shapes the formation of future

cooperative ties. This process thus prompts an endogenous dynamic that drives the emergence of
a set of opportunities and constraints that eventually circumscribe the landscape of corporate
competition (see also Walker, Kogut and Shan 1997).

Another thread running through many of the studies is the theme of external legitimacy

and status imperative imposing upon alliance dynamics. How firms ally themselves to others not
only influences its capabilities but also how others perceive its capabilities (Stuart et al., 1999;
Calabrese, Baum and Silverman 2000). Firms that have well-known affiliates enjoy a significant
advantage in contests for the recognition and acceptance of their products. Thus, if a new firm
lacks resources and suffers in the marketplace from uncertainty, and if alliances provide favorable
signals about the firm when its true qualities are least well know, then ties to prominent market
actors or legitimate institutions should provide a significant buffer against the hazards typically
faced by start up companies. In the high-tech industry, this certification advantage enables firms
that have higher level of innovative capabilities to attract more exchange partners (Stuart 1998),
and help those allied to large and innovative partners perform better than otherwise comparable
firms that lack such allies (Stuart et al, 1999; Calabrese, Baum and Silverman 2000; Powell et al.,
1996)

The primary premise inherent to those sociological studies is that organizations in general

have difficulties in obtaining information about the competency and reliability of the potential
partners. Network ties serve one way or another to ameliorate such difficulties commonly faced by
market actors. All those accounts have advanced our understanding of alliance behavior and
invigorate the embeddedness perspective in the studies of corporate activities; they have, however,
paid less attention to the broader institutional configuration and examine how alliance activities
are constrained within its operation. Although empirical studies have demonstrated that
institutional linkages –defined as ties to prominent governance agencies or community
organizations with administrative authority— do exhibit significant survival advantages over
increasing environmental competition (Baum & Oliver 1991), how this linkage advantage may
influence the formation of alliance is a question yet to be fully answered. Reviewers on strategic

alliances have recently called for a political-economic framework that takes into account the
institutional factor for better understanding alliances in divergent economic circumstances

(Todeva and Knoke 2002). Studies of this sort have only begun to scratch the surface (Hitt et. al.,
2004).

In this study, I intend to take on this relatively unexplored topic and examine how

corporate alliances are influenced by the factor of institutional embeddedness. Specifically, I ask:
Does institutional embeddedness deliver endorsement benefits for firms to form alliance networks?
Furthermore, I intend to inquire into its limit and its contingent effects of how it interacts with
social-structural determinants in jointly affecting the dynamics of alliance activities. By following
the premise that organizations tend to act to exploit the institutional context in which they are
embedded so as to stabilize the competition they face (Fligstein, 1996; Calabrese, Baum and
Silverman 2000), I hypothesize that, in a market transition period when opportunities are vaulted
over massive uncertainties and risks, institutional linkage becomes a pivotal criterion through
which market actors seek reliable and endorsed cooperative partners. The principal dimension of
this study is the selection of partners for corporate alliances. Key questions this study tries to
answer are: To what extent will institutional linkage determine which firms enter into alliances?
And, while the market is generally considered to be socially, hierarchically, and categorically
structured, to what extent will the institutional embeddedness manifest its impact upon these
structural forces?

THEORY & HYPOTHESES

Institutional Embeddedness and Formation of Alliance Networks

A general consensus is that institutions confer endorsement benefits for organizations. For

example, institutionalists have long argued that an organization is more likely to survive if it
obtains legitimacy and approval from external constituents of its institutional environment (Meyer

and Rowan, 1977; DiMaggio and Powell, 1983). Conformity to the norms and social expectations of
the institutional environment are expected to improve an organization’s life chances, and as an
institutional linkage to well-regarded institutions signals its adherence to institutional

prescriptions, institutional embeddedness enhances the level of attractiveness of focal firm
surrounded by complex webs of interorganizational exchanges.

This should especially be the case when the organizational field has undergone drastic

transition in which “cultural templates” for appropriate conduct is yet to be attained and when an
enormous amount of intensity for definitional battle is yet to come to a peace (Fligstein 1996). In
markets where opportunistic behaviors are common and accurate assessment of feasible partners
is difficult, institutional linkage may thus help a firm achieve an image of reliability and
accountability, as endorsing institutions themselves usually represent as a kind of public trust. In a
competitive environment where reliable and accountable organizations are always favored,
external access for legitimacy therefore becomes a critical selection criteria for firms to adapt into
an organizational population where institutional linkage is rare and valuable (Hannan and
Freeman 1989).

Institutional embeddedness also facilitates resource acquisition (DiMaggio 1983). Type of

resources include material such as subsidized grants from government agencies (Wiewel and
Hunter, 1985), brokered technology in educational or research institutions (Ansell 2000; Hage and
Alter 1997), or information benefit that is critical for firms to develop under regulated economies
(Schneider and Maxfield 1997, p. 7-11). Those factors, that institutional embeddedness confers
legitimate endorsement and material resources, can be powerful enabling conditions that enhance
the likelihood of focal firms to engage in alliance activities. Thus,

Hypothesis 1: The greater the extent of a firm’s institutional embeddedness, the greater the likelihood
that it will enter alliance networks.

Yet, the effect of the institutional embeddedness upon the formation of alliance network is

probably not a simple linearity in its nature. As legitimacy of a governing power is itself a variable
(Evans and Rauch 1999), over-commitment to certain firms may hinder the perceived impartiality
of institutional authority. Gould and his colleague have indicated, within the modern pluralist

polities and liberal economies, public institutions are normally defined as serving the public
interest, and thus secure their own prestige by not sliding into a stance that may favor any biased
interests (Fernandez and Gould 1994). When the connections between government agencies and

private firms are too close, there will therefore be operation of interest exchange generated in a
clientelist fashion (Evans 1995; Chibber 2002), in that legitimation effect slips and prestige tumbles.

Although connections to legitimate institutions may bring up endorsement benefits for firms to
attract alliance partners, over-embeddedness to institutional power may breed the opposite. This
reasoning should thus lead us to expect that,

Hypothesis 2: There is an inverted U-shaped relationship between the extent of institutional
embeddedness and the likelihood for a firm to enter alliance networks.

Institutional Embeddedness, Market Structure & Alliances

A common theme among organizational studies of market is stratification and

concentration (Stinchcombe 1965; White and Eccles 1986). To untangle the processes and
interorganizational mechanisms underlying stratified market order and concentration is thus a
major enterprise for contemporary organizational sociology (Stinchcombe 1998). As the market is
socially

and hierarchically structured, organizations could stratify along several dimensions such as

material resources (Pfeffer and Salancik 1978 ), age (Hannan and Freeman 1989), prestige (Podolny
1993), or experience (Gulati 1995; 1999). Studies on alliance activities have showed that higher
status firms in general have a stronger proclivity to enter alliances (Powell and Brantley, 1992;
Gulati 1999), or are more likely to attract alliance partners (Stuart 1999). Resourceful organizations
not only receive more cooperative affiliation, but also would be able to replicate their influence

upon stratified order through channels such as alliance networks. This leads stratification theories
in convergence in their consensus that the stratified order is not only stable, unlikely to change
drastically, but also self-enforcing; and it is interorganizational affiliations that tend to work as a
filter that replicates this self-enforcing status ordering (Benjamin and Podolny 1999).

Given the fact that institutional embeddedness and stratified order are equally effective in

directing the occurrences of alliances, it is less clear how these two forces interact with each other.
For example, would higher status organizations tend to gain additional advantage when they are
also endorsed by the legitimate institutions? A reasonable conjecture may lead to a positive
answer, and this can be further elaborated on and examined with its underlying mechanisms.

For institutionalists, market organizations are embedded within a larger group of

organizations -- a so called “organizational field”, in which it is the more powerful members who
benefit most from rule-setting and norm enforcement for market competition. Larger firms
provide “success stories” for other organizations to follow and this mimic isomorphism is
especially critical to institutionalize a stable market when uncertainty is high and definition of any
given situation is open to a variety of interpretations (Fligstein 1990: 7). While an institutional
authority, i.e. the state power, is considered as the most powerful contestant within this field, and
also the central arbiter of the legal and the illegal in market behavior, institutionalists have argued
that the larger firms will in general benefit from the state’s services (ibid 1990:9). This is because,
for the most part, states depend on large and successful businesses for their legitimacy. From an
institutionalist point of view, a stratified market indicates order and control, which is not only the
natural market tendency (Abolafia 1984) but is also the preferred— since governing the market as a
stabilized order indicates the efficiency of the ruling regime and hence also further reinforces its
legitimacy. Ensuring a stratified market order is therefore beneficial to both parties. This mutual
interdependence implies that in the long run the state should favor high status organizations for
the conferral of legitimacy. Bigger and more successful organizations tend to gain more favor from
legitimate authority. And the outcome is usually the preservation and reinforcement of a
dominant structure that has already prevailed within an organizational field (DiMaggio 1983: 154):

Organizations that receive grants are perceived widely as successful. Through a kind of Matthew

Effect, their managers and staff are chosen for a range of honorific positions – seats on state and

federal grant review panels, offices in trade associations, speaking engagements and

consultantships – that increase their visibility and centrality yet further. As the information load

increases, these central persons acquire power over decisions affecting the field, and a stable and

consequential dominance hierarchy emerges. [my emphasis]

This reinforcing picture can be further elaborated by identifying another underlying

mechanism that organized affiliation with the state is also self-selected: this is due to an important
constraint on collective action in which lobbying for access to the legitimate authority is feasible
only for organizations that are bigger in size and smaller in number (Olson 1965). The selective
nature of institutional endorsement could hence become a sign of an attractive value for market

actors who are seeking partners to establish alliances. This then generates an iterative dynamic: In
the long run, those who are already successful not only gain institutional linkage more easily than

others, but are also in a more favorable position to choose their alliance partners for their own
gains. This, in return, consolidates their already prominent status even more. This synthesis
should then lead us to expect that under a full-fledged market condition where the stratified order
is at work, the state should also positively interact with this stratified market force in affecting
inter-organizational activities in one way or another. On the organizational actor’s level, affiliates
of prominent entities – be they higher status firms or institutional authority-- tend to enhance the
market perceptions of focal actors as legitimate and valuable, and when institutional and market
status endorsements are simultaneously occupying their strategic vision, consolidating the
advantages gained in two sorts shall thus be a reasonable option. Thus,

Hypothesis 3: Institutional embeddedness has a stronger positive effect on alliance formation for a
firm that has higher market status than otherwise.

Institutional Embeddedness, Organizational Identity & Alliance

Organizational sociologists have only recently begun to investigate how market actors are

structured and constrained by categorical imperative in which economic exchanges are
circumscribed (Zuckerman 1999; Zuckerman and Kim 2003). Products connect to certain brand
names as firms position themselves into industries, sectors or market categories to which they
consider they belong. In such fashion, market conditions certain role-sets for firms or products as
society does for individuals (White 1981). On this market interface, positioning within
recognizable categories enables firms to appeal to audience expectation; on the contrary,
unclassifiable actors and objects may suffer social penalties because they threaten reigning
interpretive frameworks (Zuckerman 1999: 1399). Corporate identity is thus constructed through
its embedded interpretive relationships; when its identity is diffusive and vague – i.e. attached to

market categories that are hard for assumed audience to perceive -- a firm is likely to endanger
itself as illegitimate, and a discount in values may follow. For example, it is shown that the stock
price of an American firm was discounted to the extent that the firm was not covered by the
securities analysts who specialized in its industries (Zukerman 1999). On the contrary, a film will

attract a larger audience and is more likely to achieve box-office success when critics specialized in
major releases review the film and certify it as fit for the mass market (Zuckerman and Kim, 2003).

Categorical imperative may associate to alliance activities as interorganizational affiliations

usually define how a focal firm is recognized (Benjamin and Podolny 1999). With whom a focal
firm allied signals its position or role in the market. Unable to express certain alliance patterns that
are accepted as legitimate may fail to secure its position. For example, when changing institutional
beliefs predisposed firms abandon unrelated diversified investments as dominant model for rapid
growth (Davis et al., 1994; Zuckerman 2000), allying to partners in unrelated industries might
confuse interested audiences. In a time when “refocused” or “core competencies” are favored by
reigning business rhetoric, firms that straddle borders in their alliance scope may suffer the
illegitimate discounts and attract less alliance partners accordingly. Although the breadth of a
firm’s alliances is likely to affect its level of network resources (Gulati 1999), over-connectedness to
too many firms in unrelated industries may endanger focal firm’s position as illegitimate, unless
they have first secured achieved sufficient recognition to obtain success in a particular category
(Zuckerman et al., 2003).

Firms’ identities are thus structured along categories by which their alliance partners are

attached; and the more focused of the alliances in specific categories, the more distinctively they
can be discerned, therefore, the better they can be rewarded. Although this reward function may
distribute according to the extent to which firms are distinctively identified, institutional
embeddedness may, however, have differential impact along the axis of this distribution. Being
tied to legitimate institutions may effectively “lift up” firms suffering in divergent and ambiguous
identities, whereas such institutional attachment will probably add only a trifling advantage to
firms that are already typecast in recognizable market classifications. This conjecture is based on a
nuanced discernment that institutional embeddedness and typecasting probably designate two
different kinds of legitimation principle: While institutions may confer endorsement generally
effective across boundaries and in a more universalistic setting of the “field” (DiMaggio and
Powell 1983), typecasting mostly generates a converging recognition and denotes a more specific
kind of legitimacy (Zuckerman 1999: 1400). The former benefits those who survive over a wider
range of niches, while the latter help firms that appeal to a more specialized audience. While

This fad of intercorporate alliances induced by massive institutional change is paralleled by

prominent momentum of market stratification during the period. The weight of the business

groups in Taiwan's economy has been increasing since the liberalization of late 1980s, and in the
newly liberalized markets, most of the entrants have been the group subsidiaries that equip
themselves with incumbent advantages. According to Chu and Hong’s study (2002), more than a
quarter of those business groups switched their core business out of traditional sectors into growth
industries during the decade. Market liberalization has provided opportunities for business
groups to expand, and help to raise the aggregate level of market concentration.

This emerging fad of intercorporate alliance activities within business community may be

contrasted with the dominant imagery of state power upon the overall intercorporate networks.
Within the overall complex of intercorporate networks, in Taiwan, State and Party Owned
Enterprises (SPOE) not only appeared as the most prominent actors (by cohesively tying to each
other via interlocking directorates and equity holding), but also as the most frequent participants
in clique formation among the business community (Lee 1998). This begets the further question of
what kind of roles did the institutional power play during the decade phase of alliances generated
by drastic economic and political transformation? This specific historical context thus constitutes
the empirical setting upon which my questions drawn to answer.

DATA & METHOD

Sample

This study examined the factors affecting the likelihood that firms will enter equity

alliances during each period of observation. The unit of analysis is the firm-year. Instead of
focusing on strategic alliances that are based usually on contractual agreements, I center my analysis
on equity alliances in which a closer commitment for resource exchange is required. The greater the

hazards associated with an alliance, the more likely it will be equity based (Gulati 1995a). There are
usually two forms of equity collaboration: the classic joint venture and the direct minority equity
participation. Either way, an initial investment and shared equity participation is necessary, and
thus distinguishes both equity forms of collaboration from pure contractual arrangements (Pisano

1989). To test the hypotheses, a database that documents the equity alliance histories of firms was
compiled. The sample included a panel of firms during 1990 to 2003, a period that saw an

unprecedented surge in intercorporate alliances in Taiwan. For each firm-year, I constructed
records that document the number of equity participation (either a direct equity investment or
joint equity venture) that the focal firm involved. Considering the ready access for important
financial information this study demands, only firms listed in the Taiwan Security Exchange Market
during the research period are included. Excluding the firms that have incomplete information on
key variables, there are totally 211 firms included in the sample, across manufacturing, electronic,
service and financial sectors. The resulting data structure is best characterized as a cross-sectional
time-series panel. This panel design has thus allowed me to assess the structural dynamics and
contingent market factors underlying alliance formation. For each firm, I collected financial data
for each year. The data was compiled mainly through the Taiwan Economic News, a local
information service vendor, and supplemented by the annual financial disclosure report of each
company.

In order to construct variables that indicate how private firms are interconnected to state

owned sectors and government agencies, I compiled an additional panel database for state owned
companies and banks. It included eight major state-owned banks and nine state-owned companies
across the same period of time. For these seventeen state-owned firms, mainly information on
equity participation and board members was collected, in which network relations between
private and state-owned firms can thereafter be identified. The data on state-owned companies
was obtained with the help from the Commission of National Corporations, the Ministry of Economic

Affairs

, and the Ministry of Finance. Information was input manually and supplemented through

the Annual Report on State-owned Enterprise published by the Commission of National Corporations.

Variables

The dependent variable is formation of equity alliances, Alliance, which calculates the

number of equity participation that the focal firm is involved with other firms in the sample in a
given year. For institutional embeddedness, I used multiple measurements for various networks

between private firms and state agencies, including the number of equity ties that private firms are

held by state-owned business (State Equity), and board members that are commissioned from the
government (State Director). I also constructed a variable which indicated the number of a private

firm’s board members sitting in state-owned business and compared with other two. Variables of
institutional embeddedness are lagged for one year. For testing hypothesis 2, I also created
quadratic terms for each variable. As scholars tend to use organizational size and age as proxies for
organizational status, I thus followed this practice and included them to gauge how alliances are
influenced by the stratified momentum of the market. Size is log of total sales of the firm in the
previous year and Age calculates the total number of years from the time of firm’s establishment

up to the prior year. Firms may also be stratified by their experience in alliance activities. As
recurrent alliance networks that a firm accumulates from the past is demonstrated as critical
factors for further alliance activities (Gulati 1995a, 1995b; Gulati and Gargiulo 1999), I thus create
Experience

measuring the cumulative total of alliances (new or old) the firm has entered until the

previous year. For testing hypothesis 3, I created additional interaction terms between variables of
institutional embeddedness, and stratification variables, i.e. size, age, and experience. The
interaction terms are created only for those whose main variables are significant. And I reported
only the results that are significant.

The previous discussion indicates that a firm’s alliance activities are influenced by its

identity embedded within the overall alliances networks surrounding the focal firm. The identity
is constructed through examination of alliance patterns that focal firms acquire in relation to other
firms that are assigned in different industrial categories. To compute the identity measures, I
aggregate the alliance number from firm level up to industrial category level, and construct a
matrix that documents the alliance affiliations among industrial categories for each year. This is a
20 by 20 matrix as there are totally 20 industrial categories. Information on industrial categories
was consulted with the Taiwan Security Exchange Market, in which each listed company is assigned
with one specific security category. Within each matrix, the number in each cell indicates the
frequency to which two industrial categories are affiliated through equity linkages among firms. I
then calculate the network centrality score for each matrix, using the Bonacich (1987) eigenvector
measure of centrality that captures the position or role of an agent in a relational network (Podolny
1993). As the unit of the matrix analysis is on the level of industrial category, the mea sure means
the most central industrial categories are those linked to many categories through alliances, which

are in turn linked to several other sectoral categories. As this measure reveals how diffusive each
industrial sector is in its alliances across industrial boundaries, I then multiplied -1 for this
measurement to indicate how focused each industrial category is in its alliance pattern, and then
export this measure, termed as Alliance Identity, back to each firm in each year. This variable is also
lagged for one year. For testing hypothesis 5, I also created interaction variables between alliance
identity and institutional embeddedness.

Control Variables

Three control variables are included as they are expected to affect the alliance activities but

not included in the discussion of the hypotheses. While the primary focus of this research is the
role of institutional embeddedness and sociological factors in influencing alliance activities, there
are important financial and organizational attributes that could affect a firm’s alliances as well.
Business group

is a dummy variable to control the fact that firms in business group are indicated

more likely to form ties (Gerlach 1992; Granovetter 1994) and is coded one if the firm belongs to a
certain business group and zero otherwise. Several studies suggest that financially constrained or
poorly performing firms may have formed alliances because they require access to capital
resources (Gulati 1995; Stuart 1998; Mizruchi 1996). I hence included two financial variables to
control for this tendency. Profit measures the return on asset ratio (ROA). Liability Ratio is the total
liabilities divided by total assets of the firm. Variables of financial attributes for each firm are
lagged for one year.

Estimation

As the dependent variable, equity alliances, is a count variable and takes only non-negative

integer values, I estimate the alliance formation with random-effects Poisson models that is
usually adopted for panel data (Hausman, Hall, and Grilliches, 1984). I also checked on how my
definition of the firms at risk of entering equity alliances, or the risk set, might produce unbiased
results. While including many firms that never enter an alliance could lead to its own set of biases,

there were no observable criteria I could apply a priori to determine which firms were likely to
enter alliances and which were not (see also Gulati 1995: 630). As zero-inflated observations on the
dependent variable may cripple the efficiency of Poisson models estimation, I compared the

results with a more restrictively defined risk set that excludes the firms that have never entered
even one alliance during the observation period. Approximately only three percent of the firms

were excluded from such criteria and the results yielded no major difference. I thus stuck on the
original risk set and reported results from Poisson models. In the estimation, I also take into
account the issue of endogenous occurrence dependence that is normally faced by panel
estimation (Heckman and Borjas, 1980; Gulati 1999; Stuart 1998). Thus, in each estimated model I
included the variable Experience, cumulative total of alliances the firm has entered until the
previous year, as a baseline variable that controls unobserved heterogeneity in affecting the
alliance formation.

RESULTS

Table 2 reports the means and a correlation matrix for the variables in the models.

Although the sample represents the firms that have a longer establishment in their operations (my
longitudinal database requires firms to be listed on the exchange market at or before 1990), there
exists, however, considerable variance on key variables, such as age, profit, board and equity ties
with state business as well as their experience in alliance activities. Past experience in alliances is
highly correlated with formation of future alliance networks, as would be expected. Firms that are
older in age and bigger in size are also more likely to form ties, as well as to accumulate experience.
There is a fair amount of positive correlation between equity ties with state and firm size, implying
that bigger firms are also more likely to form equity ties with state owned business.

[ Tables 1 & 2 are about here ]

Table 3 reports the estimates from the random effects Poisson regressions of the equity

alliances. Model 1 presen ts the base model with only the control variables. Models 2 to 5 add
variables of institutional embeddedness and their respective quadratic terms to the specification.
Model 6 to 8 add interaction terms between institutional embeddedness and market stratification
along three dimensions, i.e., size, experience and age. Models 9 and 10 add alliance identity and its

interaction term with institutional embeddedness.

[ Table 3 is about here ]

Models 2 and 4, which add to the baseline model the institutional embeddedness, support

hypothesis 1: parameter estimates for institutional embeddedness are significantly positive,
showing that increases in equity or board ties with state multiplied the rate of alliance formation.
According to the parameter estimates, holding all other variables constant, a one standard
deviation increase in state equity multiplied the rate by a factor of 1.04 (= exp

.092*.48

), as similar to

the effect of board ties to state (1.04=exp

.053*.78

). Hypothesis 2 predicted that institutional

embeddedness has nonlinear effects on an organization’s alliance formation rates. In particular,
the argument predicts that firms will experience diminishing benefits as the alliance formation rate
will increase at a decreasing rate with upward changes in institutional embeddedness. To test this
idea, Models 3 and 5 add the quadratic terms for institutional embeddedness. The negative
parameter estimates on these variables supports the hypothesis, although only the quadratic term
for state equity is significant.

Hypothesis 3 predicted that because institutional legitimacy tends to work along with

market stratified momentum, institutional embeddedness shall fare better on alliance formation
for organizations that are high in status. The hypothesis was tested with three interaction variables
between institutional embeddedness and market stratification, i.e., size, experience in alliances
and age. Models 6 to 8 report the results. Although the parameters show expected positive and
significant effects, and thus support for the hypothesis, the evidence is marginal, or fair at best.
The interaction term between state equity and size is not significant, and the significant effect of the
interaction term between state equity and experience is negligible. The interaction term between
state equity and age has however shown a relatively satisfactory evidence that institutional effects
do work on the reinforcing dynamics with market stratification.

The final hypotheses predicted that firms embedded within industrial sectors that have

relatively distinct identities in their alliance patterns tend to have higher number of alliance
partners, and that institutional embeddedness tends to have a stronger effect on firms that are
relatively diffusive and vague in their alliance patterns. Models 9 and 10 thus add the variables of
alliance identity and its interaction term with institutional embeddedness to test the ideas. And the
evidence supports the hypotheses. The interaction term is negative and significant, demonstratin g

that the magnitude of institutional embeddedness effect is stronger for firms with lower alliance
identity. Firms that are too diffusive in their alliances, and thus suffer from a discounted

legitimacy for attracting further alliance partners, may enjoy the “lifting-up” from their ties to
institutional authority for mending their hampered identity.

CONCLUSION

My purpose in this paper has been to demonstrate how the cooperation and competition of

firms are determined by institutional components along with social-structural determinants. More
specifically, the study shows that institutional embeddedness does deliver endorsement benefits
for firms to form alliance networks. Consistent with embeddedness and social capital arguments,
institutional embeddedness can be considered as a type of network resource that increases firms’
visibility and legitimacy, and therefore enables firms to attract more partners and secure
opportunities that further expand its operational scope. The evidence demonstrates the resilience
of institutional power within the domain of market operations. In a market transition period when
opportunities are vaulted over massive uncertainties and risks, institutional linkage become one
pivotal criterion through which market actors seek reliable and endorsed cooperative partners.
Scholars have long called for studies on how state power may influence the operation of
intercorporate governance and alliances. As studies of this sort have yet to come, this paper shall
thus deliver a timely response to this unanswered call (Hage and Alter 1997; Todeva and Knoke
2002; see also Lie 1997; Granovetter 2002; Fligstein 2001).

As organizations tend to act to exploit the institutional context in which they are embedded

so as to stabilize the competition they face (Fligstein, 1996), incumbent firms that are equipped
well with such institutional ties naturally cope well with the changing environment. As my study
shows that most firms that have stronger ties with state are firms bigger in size and older in age,
this finding is also congruent well with the ecological theory in that it was argued larger and older
organizations tend to “develop dense webs of exchange, to affiliate with centers of power” and
thus “collective action becomes more reliable and accountable.” (Hannah and Freeman 1989: 80-1).
Possibly deemed as “structural inertia” that makes them less responsive to external changes in the
past, institutional embeddedness has became a “genetic trait” that appear to be a valid operative

Literature Cited:

Ansell, Christopher K. 2000. "The Network Polity: Regional Development in Western Europe." Governance: An International

Journal of Policy and Administration 13:303-33.

Baker, Wayne E. 1990. "Market Networks and Corporate Behavior." American Journal of Sociology 96:589-625.
Barley, Stephen R., John Freeman, and Ralph C. Hybels. 1992. "Strategic alliances in commercial biotechnology." Pp. 311-47

in Networks and Organizations: Structure, Form, and Action, edited by Nitin Nohria and Robert G. Eccles. Boston, MA:
Harvard Business School Press.

Barney, J. B. 1991. "Firm resources and sustained competitive advantage." Journal of Management 17:99-120.

Baum, Joel A. C., and Christine Oliver. 1991. "Institutional Linkages and Organizational Mortality." Administrative Science

Quarterly 36:187-218.

— . 1992. "Institutional Embeddedness and Dynamics of Organizational Populations." American Sociological Review 57:540-559.
Benjamin, Beth A., and Joel M. Podolny. 1999. "Status, Quality, and Social Order in the California Wine Industry."

Administrative Science Quarterly 44: 563-89

Bonacich, Phillip. 1987. "Power and centrality: A family of measures." American Journal of Sociology 92:1170-82.

Burt, Ronald S. 1992. Structural holes : the social structure of competition. Cambridge, Mass.: Harvard University Press.
— . 2000. "The Network Structure of Social Capital." in Research in Organizational Behavior. Greenwich, CT.: JAI Press.

Calabrese, Tony, Joel A. C. Baum, and Brian S. Silverman. 2000. "Canadian biotechnology start-ups, 1991-1997: The role of

incumbents' patents and strategic alliances in controlling competition." Social Science Research 29:503-34.

Carroll, Glenn R., Jerry Goodstein, and Antal Gyenes. 1988. "Organizations and the State: Effects of the Institutional

Environment on Agricultural Cooperatives in Hungary." Administrative Science Quarterly 33:233-56.

Chibber, Vivek. 2002. "Bureaucratic Rationality and the Developmental State." American Journal of Sociology 107:951-89.
Chu, Wan-wen, and Chia-yu Hong. 2002. "Business Groups in Taiwan's Post-Liberalization Economy." Taiwan: A Radical

Quarterly in Social Studies 47:33-79.

Davis, Gerald F., Kristina A. Diekmann, and Catherine H. Tinsley. 1994. "The Decline and Fall of the Conglomerate Firm in

the 1980s: The Deinstitutionalization of an Organizational Form." American sociological review 59:547.

DiMaggio, Paul J. 1983. "State Expansion and Organizational Fields." Pp. 147-61 in Organizational Theory and Public Policy,

edited by Richard H. Hall and Robert E. Quinn. Beverly Hills: Sage Publications.

DiMaggio, Paul J., and Walter W. Powell. 1983. "The Iron Cage Revisited: Institutional Isomorphism and Collective

Rationality in Organizational Fields." American Sociological Review . 48:147-60.

Dobbin, Frank, and Timothy J. Dowd. 1997. "How Policy Shapes Competition: Early Railroad Foundings in Massachusetts."

Administrative Science Quarterly 42:501-29.

Evans, Peter B. 1995. Embedded Autonomy. Princeton: Princeton University Press.
Evans, Peter B., and James E. Rauch. 1999. "Bureaucracy and Growth: A Cross-National Analysis of the Effects of

"Weberian" State Structures on Economic Growth." American Sociological Review 64:748-65.

Fernandez, Roberto M., and Roger V. Gould. 1994. "A Dilemma of State Power: Brokerage and Influence in the National

Health Policy Domain." American Journal of Sociology 99:1455-91.

Fligstein, Neil. 1996. "Markets as Politics: A Political-Cultural Approach to Market Institutions." American Sociological Review 61:

656-673

Gerlach, Michael L. 1992. Alliance Capitalism: The Social Organization of Japanese Business. Berkeley, CA: University of California

Press.

Gomes-Casseres, Benjamin. 1996. Alliance Revolution. Cambridge: Harvard University Press.
Granovetter, Mark S. 1985. "Economic Action and Social Structure: The Problem of Embeddedness." American Journal of

Sociology:481-510.

— . 1994. "Business Groups." Pp. 453 -75 in The Handbook o f Economic Sociology, edited by Neil J. Smelser and Richard Swedberg.

Princeton: Princeton University Press.

— . 2002. A Theoretical Agenda for Economic Sociology. New York: Russell Sage.

— . 2003. "Business Groups and Social Organization." in Handbook of Economic Sociology, 2nd edition.
Guillen, Mauro F. 2000. "Corporate Governance and Globalization: Arguments and Evidence against Convergence."

Wharton School, University of Pennsylvania.

Gulati, Ranjay. 1995a. "Social Structure and Alliance Formation Patterns: A Longitudinal Analysis." Administrative Science

Quarterly 40:619-52.

— . 1995b. "Familiarity breeds trust? The implications of repeated ties on contractual choice in alliances." Academy of

Management Journal 38:85-112.

— . 1999. "Network location and learning: The influence of network resources and firm capabilities on alliance formation."

Strategic Management Journal 20:397-420.

Gulati, Ranjay, and Martin Gargiulo. 1999. "Where Do Interorganizational Networks Come From?" American Journal of

Sociology:1439-1493.

Hage, Jerald, and Catherine Alter. 1997. "A Typology of Interorganizational Relationships and Networks." Pp. 94-126 in

Contemporary Capitalism: The Embeddedness of Institutions, edited by J. Rogers Hollingsworth and Robert Boyer.
Cambridge: Cambridge University Press.

Hannan, Michael T., and John Freeman. 1989. Organizational ecology. Cambridge, Mass.: Harvard University Press.

Hausman, Jerry, Bronwyn H. Hall, and Zvi Griliches. 1984. "Econometric models for count data with an application to the

patents-R&D relationship." Econometrica 52:909-38.

Heckman, James, and George J. Borjas. 1980. "Does unemployment cause future unemployment: Definitions, questions, and

answers from a continuous time model of heterogeneity and state dependence." Econometrica 47.

Higgins, Monica C., and Ranjay Gulati. 2003. "Getting off to a good start: The effects of upper echelon affiliations on

underwriter prestige." Organization Science 14:244-63.

Hitt, Michael A. , David Ahlstrom, M. Tina Dacin, Edward Levitas, and Lilia Svobodina. 2004. "The Institutional Effects

on Strategic Alliance Partner Selection in Transition Economies: China vs. Russia." Organization Science 15:173-185.

Kogut, Bruce. 1988. "Joint Ventures: Theoretical and Empirical Perspectives." Strategic Management Journal 9:319-332.
Koka, Balaji R., and John E. Prescott. 2002. "Strategic alliances as social capital: A multidimensional view." Strategic

Management Journal 23:795-816.

Lee, Zong-rong. 1998. "Business Embeddedness of the State in Taiwan." in 5th Annual Meeting for North American Taiwan

Studies Conference. WI: Madison.

— . 2003. "State Embeddedness, Market Stratification, and Inter-Corporate Alliances." in Workshop on East Asian Studies.

University of Chicago.

Lie, John. 1997. "Sociology of Markets." Annual Review of Sociology 23:341-60.
Mahoney, J. T., and J. R. Pandian. 1992. "The resource-based view within the conversation of strategic management." Strategic

Management Journal 13:363-80.

Mariolis, Peter. 1975. "Interlocking Directorates and Control of Corporations." Social Science Quarterly 56:425-39.

Meyer, John W., and Brian Rowan. 1977. "Institutional organizations: Formal structure as myth and ceremony." American

Journal of Sociology 83:340-63.

Mizruchi, Mark S. 1992. The Structure of Corporate Political Action. Cambridge: Harvard University Press.
— . 1996. "What Do Interlocks Do? An Analysis, Critique, and Assessment of Research on Interlocking Directorates."

Annual Review of Sociology:271-98.

Olson, Mancur. 1965. The logic of collective action ; public goods and the theory of groups. Cambridge,: Harvard University Press.

Padgett, John F., and Christopher K. Ansell. 1993. "Robust Action and the Rise of the Medici, 1400-1434." American Journal

of Sociology 98:1259-1319.

Pfeffer, Jeffrey , and Gerald R. Salancik. 1978. The External Control of Organizations: A Resource Dependence Perspective. New York:

Harper & Row.

Pisano, Gary P. 1989. "Using equity participation to support exchange: Evidence from the biotechnology industry." Journal of

Law, Economics, and Organization 5:109-26.

Podolny, Joel M. 1993. "A Status-based Model of Market Competition." American Journal of Sociology 98:829-72.

Powell, Walter W., K. W. Koput, and L. Smith-Doerr. 1996. "Interorganizational collaboration and the locus of innovation:

Networks of learning in biotechnology." Administrative Science Quarterly 41:116-45.

Powell, Walter W., K. W. Koput, L. Smith-Doerr, and J. Owen-Smith. 1999. "Network Position and Firm Performance:

Organizational Returns to Collaboration in the Biotechnology Industry." Research in the Sociology of Organizations

16:129-59.

Schneider, Ben Ross, and Sylvia Maxfield. 1997. "Business, the State, and Economic Performance in Developing Countries."

Pp. 3-35 in Business and The State in Developing Countries, edited by Sylvia Maxfield and Ben Ross Schneider. Ithaca:
Cornell University Press.

Stinchcombe, Arthur L. 1965. "Social Structure and Organizations." Pp. 142-93 in Handbook of Organizations, edited by James

G. March. Chicago: Rand McNally.

— . 1998. "Monopolistic competition as a mechanism: Corporations, universities, and nation-states in competitive fields." Pp.

267-305 in Social Mechanisms: An Analytical Approach to Social Theory , edited by Peter Hedstrom and Richard Swedberg.

Cambridge: Cambridge University Press.

Stuart, Toby E. 1998. "Network Positions and Propensities to Collaborate: An Investigation of Strategic Alliance Formation

in a High-technology Industry." Administrative Science Quarterly 43:668-98.

Stuart, Toby E., Ha Hoang, and Ralph C. Hybels. 1999. "Interorganizational Endorsements and the Performance of

Entrepreneurial Ventures." Administrative science quarterly 44:315 (35 pages).

Todeva, Emanuela , and David Knoke. 2002. "Strategic Alliances and Corporate Social Capital." Koelner Zeitschrift fuer Soziologie

und Soziapsychologie.

Uzzi, Brian. 1997. "Social Structure and Competition in Interfirm Networks: The Paradox of Embeddedness." Administrative

Table 1. Definition and predicted signs of variables

Variable

Definition

Alliance

Number of equity alliances focal firm involved in a given year

State Equity

Equity ties that firms are held by state-owned business

State Director

Number of board members commissioned from the government

Size

Log of total sales in previous year

Age

Number of years from the time of firm’s establishment up to the prior year +

Profit

Return on assets (ROA) in previous year

Liability

Total liabilities divided by total assets of the firm in previous year

Business group

Dummy variable to one if the firm belong to a business group

Experience

Total alliances the firm has entered until the previous year

Identity

Alliance identity using Bonacich’s centrality measure on industrial level

Table 2. Descriptive Statistics

Mean

S.D.

Lowest

Highest

(1)

Alliance

20.21

18.89

127

(2)

State Equity

0.13

0.48

(3)

State Director

0.25

0.78

(4)

Size

15.19

1.30

8.89

19.93

(5)

Age

29.72

10.44

(6)

Profit

6.43

8.17

-41.84

63.60

(7)

Liability

43.08

35.13

4.85

1163.31

(8)

Business Group

0.89

0.32

(9)

Experience

134.08

150.47

1251

(10)

Identity

31.56

5.31

3.47

39.67

Table 2.

(continued)

Correlation Matrix

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(1)

1.00

(2)

0.14

1.00

(3)

0.14

0.37

1.00

(4)

0.39

0.29

0.19

1.00

(5)

0.45

0.14

-0.04

0.18

1.00

(6)

-0.10

0.04

0.03

0.18

-0.33

1.00

(7)

0.17

0.08

0.05

0.08

0.17

-0.29

1.00

(8)

0.10

-0.02

0.06

0.16

-0.09

0.08

-0.01

1.00

(9)

0.74

0.06

0.04

0.34

0.52

-0.23

0.13

0.08

1.00

(10)

0.09

0.04

0.05

0.07

0.03

-0.04

0.04

0.01

0.09 1.00

Table 3. Maximum Likelihood, Random Effects Poisson Estimates of Equity Alliances

(1)

(2)

(3)

(4)

(5)

Constant

-.899**

-.917**

-.948**

-.933**

(.267)

Size

.252**

.251**

.252**

.253**

(.013)

Profit

-.009**

(.001)

Liability

-.005**

(.000)

(.001)

Business Group

.183

.191

.188

.176

(.195)

(.194)

(.193)

(.194)

Experience

.001**

(.000)

State Equity

.092**

.232**

(.014)

(.036)

State Board

.053**

.058**

(.011)

(.026)

State Equity²

-.049**

(.011)

State Board²

-.001

(.006)

Alpha

.767**

.763**

.764**

.760**

(.072)

2391

Log L

-10067.85

-10048.48

-10039.56

-10057.11

-10057.09

Wald Chi-square

1197.59

1234.81

1252.06

1217.66

* p< .05, ** p<.01

Table 3. (continued)

(6)

(7)

(8)

(9)

(10)

Constant

-.855**

-.924**

-1.904**

-.829**

-.806**

(.269)

(.267)

(.282)

(.269)

Size

.247**

.251**

.175**

.250**

(.013)

Profit

-.009**

-.001

-.009**

(.001)

Liability

-.005**

-.006**

-.005**

(.001)

(.000)

Business Group

.197

.195

.442

.191

.193

(.194)

(.213)

(.194)

Experience

.001**

-.001**

.001**

(.000)

State Equity

-.221

.069**

-156.**

.092**

-.057**

(.175)

(.019)

(.050)

(.015)

(.083)

State Equity * Size

.019

(.010)

State Equity * Experience

.000*

(.000)

Age

.072**

(.003)

State Equity * Age

.007**

(.001)

Identity

.003*

.003**

(.001)

State Equity * Identity

-.005*

(.002)

Alpha

.764**

.915**

.765**

.764**

(.072)

(.086)

(.072)

2391

2390

Log L

-10046.87

-10046.34

-9666.17

-10037.57

-10035.90

Wald Chi-square

1237.57

1239.65

1928.06

1217.66

1242.86

* p< .05, ** p<.01,

Graph 1. Formation of Equity Alliances in Taiwan, 1990 to 2003

Graph 2. Board Ties to Public Sector and its Impact on Alliance Formation

10 11 12

Year 1990 - 2003

Government Ties & Equity Alliances

Non Government Board (N=145)

Government Board (N=66)

Year 1990 - 2003

Formation of Equity Alliances in Taiwan

Equity Alliances