alliance portfolio diversity and firm...
TRANSCRIPT
ALLIANCE PORTFOLIO DIVERSITY AND FIRM PERFORMANCE
Ruihua Joy JiangSchool of Business Administration
Oakland University2200 N. Squirrel Rd. Rochester, MI 48309
Phone: 248-370-2832Fax: 248-370-4275
Email: [email protected]
And
Qingjiu Tom TaoCollege of Business and Economics
Lehigh University621 Taylor St. Bethlehem, PA 18015
Phone: 610-758-3438Fax: 610-758-6941
Email: [email protected]
And
Michael D. SantoroCollege of Business and Economics
Lehigh University621 Taylor Street
Bethlehem, PA 18015Phone: 610-758-6414Fax: 610-758-6941
Email: [email protected]
Running Head: Research Notes and Commentaries
Keywords: alliance portfolio diversity, partner diversity, functional diversity, governance
diversity, diversity and performance, auto industry
0
ABSTRACT
In this paper, we propose a comprehensive alliance portfolio diversity construct that include
partner, functional, and governance diversity. Ground our work primarily with the resource- and
dynamic capabilities-based views, we argue that increased diversity in partners’ industry,
organizational, and national background will incur added complexity and coordination costs as
well as broadened resource and learning benefits. Increased functional diversity results in a
balanced portfolio while increased governance diversity inhibits learning and routine building.
Hypotheses were tested with alliance portfolio and performance data for 138 multinational firms
in the global automobile industry during the twenty-year period from 1985 to 2005. We found
alliance portfolios with greater organizational and functional diversity and lower governance
diversity to be related to higher firm performance. Industry diversity had a U-shape relationship
to firm performance. We suggest that firms manage their alliance with a clear portfolio
perspective, seeking to maximize resource and learning benefits by collaborating with a variety
of organizations in various value-chain activities while minimizing managerial costs with a
focused set of organizing structures.
1
INTRODUCTION
Research on strategic alliances has proliferated commensurate with increasing use (Lavie, 2007).
The majority of the strategic alliance literature can be categorized into two main streams. The
first focuses on the alliance as the principal unit of analysis, examining determinants of alliance
formation, governance, management, and performance. The second adopts a network
perspective, exploring outcome implications of a firm’s position in a network of cooperative
relationships. While both aid our understanding of strategic alliances, recently emerging is a firm
level perspective where a firm’s alliances are viewed as a portfolio with a focus on the
composition of alliance partners (e.g., Goerzen and Beamish, 2005; Lavie, 2007; Reuer and
Ragozzino, 2006; Stuart, 2000). Our study extends this emerging literature by establishing a
comprehensive construct of alliance portfolio diversity that incorporates functional diversity (the
range of activities for which the firm uses alliances), governance diversity (the variety of
structures with which the firm manages alliances) and partner diversity (the diversity of partners
with which the firm allies), and examines these constructs with respect to firm performance.
We underpin our work with resource-based theory where the firm is a bundle of heterogeneous
resources (Barney, 1991) and the dynamic capabilities framework where competitive advantage
draws from the firm’s ability to exploit as well as combine and re-combine resources that are
valuable, rare, and hard to imitate and substitute (Teece, Pisano and Schuen, 1997). In addition to
internal operations (i.e., the make) and external acquisitions and outsourcing (i.e., the buy)
alliance strategies are often an indispensable part of a firm’s overall corporate strategy where an
effective alliance portfolio can be a vital component for exploiting resources and exploring new
opportunities for attaining superior performance.
2
THEORY AND HYPOTHESES
Firms enter alliances for many reasons such as to pool complementary resources (Eisenhardt and
Schoonhoven, 1996), share costs and risks of undertaking expensive and highly uncertain
projects (Hagedoorn, 1993), and access or acquire needed resources, capabilities or knowledge
(Powell, Koput and Smith-Doerr, 1996). Studies found that greater alliance intensity is positively
associated with firm survival (Baum and Oliver, 1991), higher firm growth rates (Powell et al.,
1996), and higher levels of innovation (Hagedoorn and Schakenraad, 1994). However,
Hagedoorn and Schakenraad (1994) found no effect of technology alliances on profitability and
Stuart’s (2000) work shows not all alliances positively impact firm performance. We go beyond
recent work on network partner attributes to examine the diversity of alliance portfolio
composition and its relationship to firm performance.
Alliance portfolio diversity
We define alliance portfolio diversity as the degree of variance in partners, functional purposes,
and governance structures of the alliances. Prior studies have focused on partner diversity at the
dyad (e.g. Parkhe, 1991) or network level (e.g. Goerzen and Beamish, 2005). We contend that
alliance portfolio diversity must also consider variance in the functional scope and governance
structure of alliances. Thus, three key issues in alliance formation can be addressed: The first
issue is partner selection; with whom the firm allies. Our notion of alliance portfolio partner
diversity is similar to Goerzen and Beamish’s (2005) network diversity and refers to the degree
of variance in partners’ resources, capabilities, knowledge, and technological bases. The second
issue is the functional purpose of an alliance; what value-chain activities the firm performs in its
alliances. The third is governance structure; how the firm organizes and manages its alliances
using different organizing structures. Alliance portfolios vary in diversity along these three
3
dimensions and despite a large alliance literature there is a lack of studies examining all three of
these alliance diversity aspects in concert. In this study, we explore these three key dimensions
of portfolio diversity and their relationship to firm performance.
Partner diversity and firm performance
Partner selection is important since accessing or learning new skills from partner(s) is a prevalent
rationale for creating an alliance (Doz and Hamel, 1998; Hagedoorn, 1993). Prior studies
examined the relationship between partner characteristics and various firm performance
measures. For instance, Hagedoorn and Schakenraad (1994) found characteristics of alliance
partners are more important than the absolute number of alliances. Stuart (2000) found that firms
achieve higher innovation and growth rates when the alliance partners are larger and possess
more technological resources, while Goerzen and Beamish (2005) found that diversity in
Japanese firms’ foreign subsidiary networks in terms of industry and country background had a
U-shape relationship with corporate performance.
Parkhe (1991) distinguished between two types of partner diversity in his discussion of inter-
firm diversity. Type I diversity forms “the underlying strategic motivations for entering into
alliances”, in that partner differences can be complementary resources that “facilitate the
formulation, development, and collaborative effectiveness” of strategic alliances. Type II
diversity refers to differences in partner characteristics that might impede communication,
encumber knowledge transfer, and increase coordination difficulty (Parkhe, 1991:580). Over
time, Type II diversities go down and Type I diversity benefits increase (Parke, 1991). At the
portfolio level firms can reap net gain benefits from increasing partner diversity when the
benefits from Type I diversity outsize the costs resulting from managing Type II diversity.
4
Firms face trade-offs as they increase the diversity of their alliance portfolio. On one hand, a
highly diversified portfolio provides broadened search options, access to enriched resource
pools, and hence added value creation and capability development opportunities. On the other
hand, increased diversity can bring more complexity, the potential for more conflicts, and hence
increased coordination and managerial costs. Thus, the arguments for and against portfolio
diversity are often equally compelling. Following studies that consider curvilinear relationships
to reconcile two compelling contrary arguments (see for example, Golden and Zajac, 2001) we
punctuate these opposing arguments and explore the notion of nonlinear relationships between
our three partner diversity dimensions and firm performance.
Industry diversity. Alliances with firms from different industries can bring a host of benefits
and costs. Partners from the same industry are often competitors, who may bring the greatest
learning through imitation and greater absorptive capacity due to an overlap in backgrounds,
experiences, knowledge and technological bases (Cohen and Levinthal, 1990). On the other
hand, conflicts of interest exist and learning races (Doz and Hamel, 1998) can happen, increasing
monitoring and safeguarding costs. Partnering in upstream or downstream activities can offer
complementary resources and/or improve value chain coordination, while partners in unrelated
industries can facilitate entry into new markets (Kogut, 1988) or stimulate technological
innovations. However, such alliances are prone to resource misfits and/or lack of synergies,
causing alliances to underperform. Partners from different industries may also have very
different routines and processes that can make collaboration difficult. Thus, while greater partner
industry diversity may provide learning and resource access benefits, firms have to first
overcome two hurdles. First, different partners often bring myriad Type II diversities that can
impede alliance value creation (e.g., conflicts with competitors, lack of synergy with partners in
5
unrelated industries). Second, increased diversity increases alliance management complexity.
These downsides appear immediately as firms start increasing the degree of partner diversity.
However, as firms become more adept at dealing with such costs and as learning and resource
benefits accumulate, they may reach a minimum degree of diversity effectiveness and can expect
net gains surpassing this threshold. Goerzen and Beamish’s (2005) work, for instance, shows a
U-shaped relationship between partner diversity (measured by industry and national background)
and firm performance. We propose:
Hypothesis 1a: Greater alliance portfolio partner industry diversity is associated with firm performance that first decreases and then increases forming a U-shape.
National diversity. Alliances increasingly involve firms from different nation states. Cross-
border alliances can facilitate market entry (Glaister and Buckley, 1996), provide complementary
capabilities (Lane, Salk, and Lyles, 2001) and enhance different knowledge bases and learning
(Lubatkin, Florin, and Lane, 2001), but can pose a high potential for conflicts. For example,
Parkhe (1991, 1993) identified and tested five dimensions of diversity that might cause tension at
the alliance level: societal culture, national context, corporate culture, strategic direction, and
management policies/organization. Cross-border partners bring differences in political economic
systems, societal and cultural institutions, government policies, and national industry structures.
Such macro-level factors often result in differences in corporate culture, strategic direction, and
management practices where diversity in these various dimensions can cause conflicts and
increase alliance costs. Tung (1993), for instance, contrasted different communication patterns
between Americans and Japanese managers and offered a typology that punctuates the
difficulties in the interaction/acculturation process involving nationals from different countries.
At the portfolio level, allying with partners from different countries provides market and
resource access advantages. Yet, high national diversity can create excessive coordination and
6
integration costs due to the complexity in managing Type II diversities, thus firms must deal with
the downside of increased national diversity before they can reap net benefits. We propose:
Hypothesis 1b: Greater alliance portfolio partner national diversity is associated with firm performance that first decreases and then increases forming a U-shape.
Organizational Diversity. More global, dynamic, and competitive environments have forced
many firms to develop core competencies in fewer areas and consequently rely on external
sources for supplementary and complementary capabilities and knowledge (Santoro and
Chakrabarti, 2002). Alliances can be formed between large multi-divisional firms, between
public and private firms, and between for-profit businesses and not-for-profit organizations such
as universities and government agencies. Different types of organizations offer different pools of
resources and capabilities (Harrison, Hitt, Hoskisson, and Ireland, 2001). For example, large
public pharmaceutical firms often ally with small private biotech firms to tap into proprietary
technologies or specific technological expertise while smaller private biotechnology firms
benefit by gaining access to financial, marketing, and/or managerial resources and capabilities
(Santoro and McGill, 2005). Not-for-profit universities and government agencies can provide
basic and pre-competitive research that complements and supplements an industrial firm’s
shorter-term and more applied R&D efforts (Santoro & Chakrabarti, 2002). Firms allying with
organizations of different sizes, structures, and purposes increase their breadth of search,
learning capabilities, resource access, and reduce the threat of core rigidities. However, diverse
organizations tend to have different goals, decision-making processes, and systems that cause
communication and coordination difficulties, resulting in increased managerial costs. We
therefore propose:
Hypothesis 1c: Greater alliance portfolio partner organizational diversity is associated with firm performance that first decreases and then increases forming a U-shape.
7
Functional diversity and firm performance
Alliances can serve different functional purposes as firms can employ marketing, manufacturing,
and distribution alliances to broaden their market reach, enhance value creation, and further
exploit core competencies (Prahalad and Hamel, 1990). R&D alliances are prevalent in many
industrial sectors as firms take advantage of rapid technological advances while dealing with
soaring costs and risks of R&D (Hagedoorn, 1993). R&D involves exploration activities while
marketing, manufacturing, and distribution involve exploitation activities. Firms must balance
both, because exploitation relates to a firm’s current viability while exploration may critically
influence a firm’s future viability (March, 1991). As a firm increases the functional diversity of
its alliance portfolio it builds a more balanced portfolio that incorporates core and non-core
activities, gains access to supplementary and complementary assets, and expands its knowledge
base as well as market reach. This balanced approach can extend the firm’s range of value
creation activities, increase flexibility, and enhance overall firm performance. Following this
reasoning, we propose:
Hypothesis 2: Greater alliance portfolio functional diversity is positively associated with firm performance.
Governance diversity and firm performance
Alliances can be structured as non-equity or various equity-ownership arrangements where
different governance structures have different implications on level of commitment, degree of
integration, and learning (Kogut, 1988). Matching structure with the characteristics of each
cooperative venture is important to balance value creation and value appropriation (Lavie, 2007)
and reduce transaction costs (Santoro and McGill, 2005). However, it takes time and repeated
iterations to understand how to set up and manage a particular governance form since each
8
governance structure requires unique resource commitments, managerial attention, and
relationship building routines. Sampson (2005) found that repeated experience with a specific
governance structure can help the firm accumulate and institutionalize knowledge and skills
about a particular governance form. Once institutionalized, this knowledge becomes
organizational routines and operational protocols that can be readily applied to future alliances,
thereby reducing managerial costs. While attempting to match governance structures to the
variety of relationships in the portfolio can be thought of as a way to deal with transaction cost
issues, firms will be better served with a focused set of familiar structures rather than trying a
new governance form with each new alliance. Excessive experimenting with different
governance structures can significantly increase managerial complexity and lost opportunities for
institutional learning, yet provide minimal additional reduction in transaction cost hazards.
Following this reasoning, we propose:
Hypothesis 3: Greater alliance portfolio governance diversity is negatively associated with firm performance.
METHOD
We used the SDC Platinum Database to identify alliance portfolios for firms in the global
automobile industry (SIC 3714, 3711, 3751, 3713, 3715) during the period January 1985 to
December 2005. Performance data were obtained from Research Insight and Thompson Banker
One. We triangulated these data and obtained data sets for 138 multinational firms. A two-period
longitudinal panel for the period 1985-2005 with two time points (2000 and 2005) were created
to examine the cumulative effect of a firm’s alliance activities on a lagged measure of firm
performance. Year 2000 and 2005 were chosen for comparison since the number of alliances in
our sample frame was skewed toward the latter years of this twenty year period.
9
Dependent variable
Corporate-level three-year average of net profit margin (2000-2002 for time point one and 2005-
2007 for time point two) is our performance measure. Net profit margin has been used in a
number of previous studies to determine firm performance (e.g., Min and Wolfinbarger, 2005)
and is useful for comparing corporate performance in the same industry. A lagged three-year
average performance in each time period was used to reduce the likelihood of spurious results
due to one unusual performance year (Tanriverdi, 2006).
Independent and control variables
Alliance Portfolio Diversity is a multi-dimensional construct that includes partner, functional,
and governance diversity. We first coded each individual alliance into different categories. For
Industry Diversity, alliances were coded into 5 categories: “4” for an alliance formed with a
partner in the same 4-digit SIC code; “3” same 3-digit SIC code; “2” same 2-digit SIC code; “1”
same 1-digit SIC code, and “0” sharing no SIC code. For Organizational Diversity, two
measures were created to capture two, but related categories. Organizational Diversity I codes
alliances for ownership: “1” for same ownership (public-public, private-private); “0” different
ownership (public-private, public-government, private-government); Organizational Diversity II
codes alliances for organizational level: “1” same organization level (parent-parent, subsidiary-
subsidiary); “0” for different Org. level (parent-subsidiary). For Functional Diversity we coded
alliances into 4 categories where marketing is “1”, manufacturing “2”, R&D “3”, and Other “4”
(other activities that had insufficient numbers to justify individual coding). For Governance
Diversity we coded each alliance into 6 categories: “1”= non-equity, “2”= minor equity up to
20%, “3”= substantial share 21-49%, “4”=equal share 50%, “5”= major share 51-79% and
“6”=dominant share 80+%.
10
We then obtained our portfolio level diversity measures by using the Blau Index of Variability
(Blau, 1977). The Blau Index has been widely used in the group diversity literature to measure
heterogeneity of categorical variables where for any given diversity variable: D= 1 - pi2 where
D represents degree of diversity, p represents the proportion belonging to a given category, and i
is the number of different categories. The variables range from 0 (a perfectly homogeneous
group) to 1 (a perfectly heterogeneous group, with members spread evenly among all categories).
Since the size of each firm’s alliance portfolio was different (ranging from 2 to 354 alliances),
we divided the absolute diversity score by the highest possible Blau score based on the size of
the firm’s alliance portfolio. For National Diversity we coded a continuous variable, with “0” for
a portfolio with alliances with no foreign partners, “1” with partners from 1 foreign country, “2”
with partners from two foreign countries, and so on.
Our single industry controls for much variance. In addition, we included three control variables,
firm size (3 year average of total assets for the years 2005, 2006 and 2007 and for the years
2000, 2001, and 2002 in Hundred Million US $), portfolio size (the firm’s total number of
alliances at year 2000 and 2005), and portfolio size squared for a potential curvilinear effect.
Models and Results
We used Random-effects GLS Regression Analysis for hypotheses testing. Random-effects GLS
is considered more efficient and appropriate than fixed effect modeling if the data can pass the
Hausman’s Test. A Hausman test showed the coefficient estimates provided by the random
effects estimator were not significantly different to those offered by the fixed effects model
(Model 2 Χ², 8d.f. = 9.42, p = 0.31 and Model 3 Χ², 12d.f. = 11.3, p = 0.50) thus the more
efficient random effects model was used. Table 1 presents the descriptive statistics and
correlations, Table 2 our GLS Regression Analysis.
11
------Insert Tables 1 and 2 Here------
Model 1 included only the control variables, Model 2 added all our portfolio diversity variables
and Model 3 the quadratic terms for partner diversity. We mean centered the variables to avoid
multi-collinearity issues arising from adding quadratic terms. The control variables show
consistent results across all models and are consistent with previous studies. Model 3 shows the
root term for Industry Diversity is significant and negative (β = -2.8, p<0.05) while the squared
term is significant and positive (β = 7.35, p<0.05) indicating a U-shape relationship and
providing support for H1a. National Diversity is not significant thus H1b is not supported. Both
Organizational Diversity root terms are positive and significantly related to firm performance (β
= 6.69, p< 0.05, and β = 11.98, p<0.001 respectively). However since the root and square terms
for our Organizational Diversity measures are both positive and significant this indicates a J-
shape relationship rather than a U-shape relationship as we proposed in H1c. Model 3 shows
functional diversity is positively related to firm performance (β = 5.29, p<0.05) and governance
diversity is negatively related to firm performance (β = -2.72, p< 0.05) offering support for
hypotheses 2 and 3. We discuss our results and implications below.
DISCUSSION AND CONCLUSION
Overall, we found a complex relationship between alliance portfolio diversity and firm
performance. Mixed results were found regarding our three partner diversity constructs and firm
performance. Partner Industry Diversity has a U-shaped relationship with firm performance,
similar to Goerzen and Beamish’s (2005) findings in their study of Japanese firms while
Organizational Diversity showed a strong positive J-shaped relationship with firm performance,
suggesting benefits from collaborating with different types of partners (e.g., private and public,
for-profit and non-profit) outweigh the costs of managing these types of relationships. Functional
12
Diversity was found desirable as a diverse portfolio of value chain activities such as alliances in
R&D, manufacturing, and marketing brings both exploration and exploitation opportunities as
firms combine and leverage complimentary and supplementary resources and capabilities.
Governance diversity was found to be negatively associated with firm performance, reflecting
the importance of accumulating knowledge through repetition for managing and organizing
cooperative relationships.
Our results highlight the trade-offs in managing a diverse alliance portfolio between broadened
search, resource, and learning benefits and increased managerial complexity and costs as a firm’s
alliance portfolio diversity increases. Greater Organizational Diversity and Functional Diversity
can be beneficial while greater Governance Diversity detrimental to firm performance. We
believe organizational and functional diversity capture the learning and resource access benefits
accruing from collaborating with diverse organizations for a wide scope of functional purposes
while governance diversity points to the organizing difficulty and complexity of dealing with
myriad structures. Our diversity measures punctuate these results since a maximum diversity
score means a firm has an equal number of alliances in each category. In the case of industry
diversity, for example, the firm with a highly balanced alliance portfolio is collaborating with a
variety of partners in different sectors that add to their knowledge base. In the case of
governance diversity it means the firm employed an equal number of various types of ownership
structures where more is not necessarily better; too many governance forms are creating undue
complexity and deterring knowledge accumulation through a lack of systematic learning. These
findings illuminate the difficult tensions that must be balanced as firms reach out to different
partners in an attempt to maximize learning and at the same time, manage costs by not losing
focus on unfamiliar and under practiced governance forms.
13
One of our key contributions is advancing our multi-dimensional alliance portfolio diversity
construct that extends prior work focused solely on partner attributes. Future work could build
upon our functional and governance diversity measures to examine how widely firms use
alliances for various learning and value-chain activities and their linkage to different governance
forms. For practicing managers, our results highlight the strategic importance of developing a
comprehensive firm-level alliance strategy, adopting a portfolio perspective, installing an
alliance portfolio management process, and actively managing such a portfolio to develop
capabilities for improving firm performance (Hoffman, 2007).
Managers should consider collaborating with different organizations to access diverse pools of
resources and knowledge and use alliances for different value chain activities to enhance value
creation and flexibility since benefits appear to emanate from learning from and working with
partners on different activities. We caution managers not to be overly zealous in allying with
partners from too many different countries nor employing many different alliance governance
forms since they may better contain costs by using a framework of focused organizing structures.
Our data shows alliances are important to the global auto industry as our sample represents a
significant portion of the global automobile industry network. Since our one industry design
limits the generalizability of our findings, future research could investigate additional industries
such as telecommunications and electronics where alliances are also prevalent. Our results show
a non-significant relationship between national diversity and firm performance. We surmise the
auto industry’s widespread use of global alliances (as shown in our data and recent conversations
one of the authors had with auto industry executives) reduced the impact of national diversity.
Although beyond the scope of this current study, this could be examined in a future study. Future
research could also build upon our multi-dimensional alliance portfolio diversity construct to
14
examine the ways firms can optimize the learning and resource benefits while limiting the
complexity and coordination costs of managing a diverse alliance portfolio.
ACKNOWLEDGEMENTS
We thank the editor and two anonymous reviewers for their constructive feedback, Professor
Ravi Parameswaran and Nandu Nayar for their help with the project, and Martindale Center at
Lehigh University for their financial support.
REFERENCES
Barney J. 1991. Firm resources and sustained competitive advantage. Journal of Management 17 (1): 771-792.
Baum J, Oliver C. 1991. Institutional linkages and organizational mortality. Administrative Science Quarterly 36(2): 187-218.
Blau PM. 1977. Inequality and Heterogeneity. New York: The Free Press. Cohen WM, Levinthal DA. 1990. Absorptive capacity: A new perspective on learning and
innovation. Administrative Science Quarterly 35(1): 128-152.Doz YL, Hamel G. 1998. Alliance advantage: The art of creating value through partnering.
Boston, MA: Harvard Business School Press.Eisenhardt KM, Schoonhoven CB. 1996. Resource-based View of Strategic Alliance Formation:
Strategic and Social Effects in Entrepreneurial Firms. Organization Science 7(2):136-150.Glaister KW, Buckley PJ. 1996. Strategic motives for international alliances formation. Journal
of Management Studies 33(3): 301-332.Goerzen A, Beamish PW. 2005. The effect of alliance network diversity on multinational
enterprise performance. Strategic Management Journal 26(4): 333-354. Golden B, Zajac EJ. 2001. When will boards influence strategy? Inclination x power = strategic
change. Strategic Management Journal, 22: 1087-1111.Hagedoorn J. 1993. Understanding the rationale of strategic technology partnering:
Interorganizational modes of cooperation and sectoral differences. Strategic Management Journal, 14(5): 371-385.
Hagedoorn J, Schakenraad J. 1994. The effect of strategic technology alliances on company performance. Strategic Management Journal 15(4): 291–309.
15
Harrison JS, Hitt MA, Hoskisson RE, Ireland RD. 2001. Resource complementary in business combinations: extending the logic to organizational alliances. Journal of Management 27(6): 679-690.
Hoffman WH. 2007. Strategies for managing a portfolio of alliances. Strategic Management Journal 28: 827-856.
Kogut, B. 1988. Joint ventures: Theoretical and empirical perspectives. Strategic Management Journal, 9(4): 319-332.
Lane PJ, Salk JE, Lyles MA. 2001. Absorptive capacity, learning, and performance in international joint ventures. Strategic Management Journal 22(12): 1139-1161.
Lavie D. 2007. Alliance portfolios and firm performance: value creation and value appropriation in the U.S. software industry. Strategic Management Journal 28: 1187-1382.
Lubatkin M, Florin J, Lane PJ. 2001. Learning together and apart: a model of reciprocal interfirm learning. Human Relations 54(10): 1353-1378.
March JG. 1991. Exploration and exploitation in organizational learning. Organization Science 2(1): 71-87.
Min S, Wolfinbarger M. 2005. Market share, profit margin, and marketing efficiency of early movers, bricks and clicks, and specialists in e-commerce. Journal of Business Research 58(8): 1030.
Parkhe A. 1991. Interfirm diversity, organizational learning, and longevity. Journal of International Business Studies 22(4): 579-601.
Parkhe A. 1993. Partner nationality and the structure-performance relationship in strategic alliances. Organization Science 4(2): 301-324.
Powell W, Koput D, Smith-Doerr J. 1996. Interorganizational collaboration and the locus of innovation: Networks of learning in biotechnology. Administrative Science Quarterly 41(1): 116-145.
Prahalad CK, Hamel G. 1990. The core competence of the corporation. Harvard Business Review 68(3): 79-91.
Reuer JJ, Ragozzino R. 2006. Agency hazards and alliance portfolios. Strategic Management Journal 27(1): 27-43.
Sampson RC. 2005. Experience effects and collaborative returns in R&D alliances. Strategic Management Journal 26(11): 1009-1031.
Santoro M, Chakrabarti A. 2002. Firm size and technology centrality in industry-university interactions. Research Policy 31(7), 1163-1180.
Santoro M, McGill J. 2005. The effect of uncertainty and asset co-specialization on governance in biotechnology alliances. Strategic Management Journal 26(13): 1261-1269.
Stuart T E. 2000. Interorganizational alliances and the performance of firms: A study of growth and innovation rates in a high-technology industry. Strategic Management Journal 21(8): 791-811.
Tanriverdi H. 2006. Performance effects of information technology synergies in multibusiness firms. MIS Quarterly 30(1): 57-77.
Teece DJ, Pisano G, Schuen A. 1997. Dynamic capabilities and strategic management. Strategic Management Journal 18(7): 509-530.
Tung RL, 1993. Managing cross-National and intra-national diversity. Human Resource Management 32(4): 461-477.
16
Table 1. Descriptive statistics
Variables Mean S.D. 1 2 3 4 5 6 7 8
Net Profit
Margin
3.11 5.05
Firm Size 23.65 54.38 .07
Portfolio
Size
19.15 39.08 -.07 .53***
Industry
Diversity
0.61 0.29 -.21*** .11† .17**
National
Diversity
5.32 6.19 -.10† .38*** .27*** .23***
Organ.
Diversity I
0.30 0.20 -.15* .01 .24*** .28*** .24**
Organ.
Diversity II
0.34 0.17 .07 .16** .18** .21*** .21** .25***
Functional
Diversity
0.60 0.26 .08 .15** .09 -.07 -.07 .24*** .25***
Gov.
Diversity
0.36 0.26 -.12* .05 .14* .02 .18** .25*** .27*** .37***
N=276, †p<0.10, *p<0.05, **p<0.01, ***p<0.001
All variables pooled for descriptive purposes.
17
Table 2. Random-effects GLS regression results
Independent variables Model 1 Model 2 Model 3
Firm size .29 .10 .22
Portfolio size -.03 -.02 .05
Portfolio size squared .00 .00 .00
Industry diversity -5.02* -2.8*
National diversity .03 -.14
Organizational diversity I
Organizational diversity II
Functional diversity
-2.78*
2.13*
4.60*
6.69*
11.98***
5.29*
Governance diversity -3.45* -2.72*
Industry diversity squared
National diversity squared
Organizational diversity I squared
Organizational diversity II squared
Model F 4.17* 37.19***
7.35*
-.01
55.32***
49.10**
83.56***
R² within .00 .00 .02
R² between .02 .23 .40
R² overall .02 .16 .29
1. Regression Coefficients (β) are shown
2. †p<0.10, *p<0.05, **p<0.01, ***p<0.001
3. Number of observations=276; Number of groups (firms) = 138
18