Case 1152821357

EMPIRICAL ESSAYS ON CORPORATE INNOVATION:
UNTANGLING THE EFFECTS OF CORPORATE VENTURE CAPITAL
by
SERGEY ANOKHIN
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
Department of Marketing and Policy Studies
Weatherhead School of Management
CASE WESTERN RESERVE UNIVERSITY
August, 2006
CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the dissertation of
SERGEY ANOKHIN
______________________________________________________
candidate for the Ph.D. degree *.
LEONARD LYNN
(signed)_______________________________________________
(chair of the committee)
WILLIAM SCHULZE
________________________________________________
SIMON PECK
________________________________________________
SHAKER ZAHRA
________________________________________________
________________________________________________
________________________________________________
6/13/06
(date) _______________________
*We also certify that written approval has been obtained for any
proprietary material contained therein.
i
This thesis is based on three appended essays. The appended essays are:
Essay 1. Corporate venture capital and internal R&D: The interplay in the
context of corporate innovation
Essay 2. Making sense of corporate venture capital II: Technology push vs.
Market pull
Essay 3. Do CVC children make better CVC parents?
ii
iii
TABLE OF CONTENTS
LIST OF TABLES 2
LIST OF FIGURES 3
ACKNOWLEDGMENTS 5
ABSTRACT 6
INTRODUCTION 9
REFERENCES 20
APPENDED ESSAYS 22
ESSAY 1. CORPORATE VENTURE CAPITAL AND INTERNAL R&D: THE INTERPLAY 24
IN THE CONTEXT OF CORPORATE INNOVATION
ESSAY 2. MAKING SENSE OF CORPORATE VENTURE CAPITAL II: TECHNOLOGY 68
PUSH VS. MARKET PULL
ESSAY 3. DO CVC CHILDREN MAKE BETTER CVC PARENTS? 109
1
LIST OF TABLES
Essay 1
Table 1. Descriptive Statistics 61
Table 2. Rates of Corporate Innovation and the CVC-R&D Interplay 62
Table 3. Industry Classification 66
Essay 2
Table 5. Models Summary 107
Essay 3
Table 7. Regression Results 140
2
LIST OF FIGURES
Essay 1
Figure 1. Intertemporal Production Frontier in Computer and Electronic Product 60
Manufacturing Industry, 1998-2003
Figure 2. Interaction of CVC, R&D and Organizational Slack 63
Essay 2
Figure 3. Production Frontier Shift in Computer and Electronic Product 105
3
4
ACKNOWLEDGEMENTS
There are a number of people who have contributed directly and indirectly to this
dissertation. First and foremost, I would like to acknowledge my dissertation committee
members: Prof. Leonard Lynn, Prof. Simon Peck, Prof. William Schulze, and Prof.
Shaker Zahra. Without their guidance and support this day would have never come.
Special thanks are in order for Prof. Robert Hisrich and Prof. Jeffrey Glass who helped
me develop my interest in corporate venture capital.
I owe a debt of gratitude to my family. First, I would like to thank my parents, Alexander
and Olga Anokhin, for instilling in me the desire to never stop learning and ultimately
contributing to my decision to pursue greater academic goals. I am especially grateful to
my wife Tanya for her patience, love, and encouragement during these years of my
studies and to my children Veronica and Matthew for keeping me motivated.
Finally, I would like to thank my fellow students – Koen Dewettinck, Danail Ivanov,
Sanjukta Kusari, Mark Meldrum, Chris Stevens, Joakim Wincent, Yasuhiro Yamakawa,
and Jun Ye – for their friendship during the PhD program.
5
Empirical Essays on Corporate Innovation:
Untangling the Effects of Corporate Venture Capital
Abstract
by
SERGEY ANOKHIN
The dissertation combines three empirical essays on corporate venture capital
(CVC) in the context of corporate innovation. It analyzes the effects of size and
composition of CVC portfolio on the rates of corporate innovation and scale efficiency
gains. It also investigates what makes some corporate parents better than others in
absorbing the know-how developed elsewhere and addresses the long term effects for the
new ventures of accepting CVC funding at founding.
The analysis demonstrates that while both CVC and R&D appear to be positively
related to the rates of corporate innovation, CVC may have a long-term effect on
innovation while R&D is most influential in the immediate perspective. CVC intensity
positively moderates the effect of R&D on innovation when organizational slack is high,
and negatively when it is low. The study confirms the presence of the inverted U-shaped
relationship between CVC intensity and corporate innovation thus suggesting existence
of the optimal size of CVC portfolio which maximizes corporate innovation.
6
Technological and market fit between the corporation and the new venture affect
the likelihood of fulfilling incumbents’ strategic objectives in different ways. Market
development potential – the ability of the startup to create new or extend existing markets
for the incumbent – is the key dimension. When market development potential (MDP) is
present, rates of corporate innovation increase and scale efficiency gains follow. When
MDP is combined with technological fit, innovativeness soars. Without MDP, however,
both rates of corporate innovation and scale efficiency decline regardless of
presence/absence of technological fit.
History matters in structural decisions of the corporations to establish CVC
programs and in their ability to realize innovative potential inherent in the new ventures’
ideas. Corporations started with the help of CVC are significantly more likely to initiate
CVC programs of their own. This is especially true for CVC children parented by the
corporations with clear policies of taking seats on the boards of their CVC children. CVC
children also make better CVC parents in terms of their ability to absorb the know-how
developed elsewhere. At the same time, they demonstrate inferior ability to utilize their
own research and development.
7
8
INTRODUCTION
Motivation for the Dissertation and Summary of the Appended Essays
Corporations are viewed as innovation engines (Schumpeter, 1942); yet
successful innovation often requires access to knowledge that is not available within the
organizational boundaries (Cohen & Levinthal, 1990). To access this knowledge
corporations design sophisticated vehicles that allow identifying and capturing external
innovative ideas. Corporate venture capital (CVC) – equity investments in
entrepreneurial ventures by incumbent firms not made solely for financial gain
(Chesbrough & Tucci, 2004; Dushnitsky & Lenox, 2005a) – appears to be an attractive
organizational arrangement allowing incumbents to tap into the know-how developed
elsewhere (Dushnitsky, 2004): over 450 leading corporations had active CVC programs
within the last decade. Thus, it should be hardly surprising that inquiry into corporate
venture capital seems to have exploded recently (Birkinshaw & Hill, 2003; Chesbrough,
2002; Chesbrough et al., 2004; Dushnitsky, 2004; Dushnitsky & Lenox, 2005b; Ernst &
Young, 2002; Gompers & Lerner, 1998; Hellmann, 1998, 2002; Keil, 2004).
At least three distinct sets of literatures have contributed to the development of
the current state of the field: venture capital, innovation management, and corporate
entrepreneurship (intrapreneurship). In particular, there is a literature that compares CVC
performance to that of independent and various forms of captive VC funds (government-
funded, finance-affiliated) (Fleming, 2004; Wang, Wang, & Lu, 2002), a literature that
examines the innovative potential of different CVC program configurations (Dushnitsky
9
et al., 2005a), and a literature that addresses how CVC programs fulfill the basic goals of
intrapreneurial development (Birkinshaw et al., 2003; Dushnitsky, 2002; Dushnitsky et
al., 2005b; Dushnitsky & Lenox, forthcoming). Finally, there is a literature that
investigates the implications of specific CVC program designs for the overall program
efficiency (Hellmann, 1998, 2002) and a rather extensive research that looks at the
corporate venturing processes from the point of view of new ventures rather than
incumbent corporations themselves (Maula, 2001; Maula, Autio, & Murray, 2003; Maula
& Murray, 2000; Zahra, 1996; Zahra & George, 1999).
Corporate venturing research to date has developed an understanding of different
aspects of corporate venturing program objectives (Christopher, 2000; Ernst & Young,
2002; Hellmann, 2002; Reaume, 2003), designs (Burgelman, 1983; Hellmann, 1998), and
success rates of CVC programs (Chesbrough, 2002; Dushnitsky et al., 2005a,
forthcoming; Ernst & Young, 2002; Gompers, 2002). Yet, despite the now voluminous
literature on corporate venturing there is still no consensus as to whether and how
corporations benefit from equity investments into new ventures; empirical results and
conceptual arguments are often contradictory or inconclusive. Researchers are still to
properly address some of the most fundamental concerns that motivate much of CVC
literature: (1) what is the optimal mixture between internally funded corporate research
and outside ideas accessed through CVC? (2) how does a composition of the incumbent’s
CVC portfolio affect the rates of corporate innovation and other strategic goals pursued
by corporations through CVC? (3) what are the long-term effects of CVC investments on
the new ventures? Does accepting CVC funding at founding affect the investees’
structural decisions beyond the timeframe of the corporate parent – new venture
10
relationship? (4) what makes some corporate parents better than others in absorbing
innovative potential inherent in new ventures’ ideas? The present dissertation takes a step
towards addressing these questions.
First essay looks at the interplay of CVC and R&D in the context of corporate
innovation and employs agency as a theoretical framework. It demonstrates that R&D
intensity is positively related to realized corporate innovation as measured by the rates of
technical change. This agrees to the results reported by the earlier studies: indeed,
commitment to R&D is necessary for innovation. It also appears that the effect of R&D
intensity on the rates of corporate innovation may be more prominent in the immediate
and not in the long-term perspective. This is in stark contrast to the effect of CVC
intensity which is a particularly significant predictor of the rates of corporate innovation
when considered with a forward lag of one to two years. This could indicate that firms
adopt the ideas generated by their own R&D function faster than those produced by the
outside ventures. If true, it may suggest the necessity of developing additional vehicles
and building absorptive capacity to accelerate adoption of the externally accessed know-
how.
This longer timeframe required for the effects of CVC to be realized by the
corporate parent may also necessitate the need for developing integrative capabilities. To
be able to internalize the know-how developed elsewhere, the firm must be able to
properly capture the external idea and to effectively combine it with the knowledge that
exists in-house. Among other things, that may require considering special governance
arrangements for the CVC unit, careful composition of the CVC unit in terms of the
functions represented in the CVC team, developing policies with respect to taking seats
11
on the boards of the new ventures supported with the corporate venture capital, and so on.
The question of leadership in terms of the organizational hierarchy is also important:
having an overly senior or an overly junior champion of the CVC unit may turn
detrimental for the success of the project.
Alternatively, much shorter time required for the effects of internal R&D to be
realized by the corporation could also indicate that internally advanced ideas are simply
focused on the short term as they need to address problems that exist here and now. If
true, it may suggest that internal research and development could develop innovation
myopia of some sort. Applied research without the fundamental component is unlikely to
sustain itself for long. In this case, if the focus of corporate R&D has indeed shifted to the
short-term perspective, the corporations may consider developing strong relationships
with the institutions that engage in fundamental research such as universities and research
labs. This should also help build external integrative capability and thus positively affect
the adoption by the incumbent corporations of the ideas developed elsewhere.
More importantly, however, it suggests that qualitatively R&D and CVC are not
perfect substitutes. To ensure successful technical advancement in short- and long-term
perspectives corporations may be better off supporting both own and outside suppliers of
innovative ideas. Thus, the very question of ‘complementary’ vs. ‘substitutive’
relationships between R&D and CVC in the context of corporate innovation may receive
a whole new meaning. The question may change from ‘whether to establish a CVC
program’ to ‘how much funds dedicate to the program’ and ‘what types of ventures to
support’. CVC investments are not homogenous; yet, little is known about the effect of
each of the different types of investments on corporate innovativeness.
12
In addition to demonstrating positive linear effects of CVC intensity on realized
corporate innovation this study confirms existence of an inverted U-shaped relationship
between the level of CVC intensity and realized corporate innovation, which makes it
possible to determine the optimal size of CVC portfolio and to balance CVC and R&D
activity of the firm in such a way as to maximize corporate innovation. The question
remains whether or not this effect is subject to some inter-firm differences that is not
captured by the study. It may be that the threshold beyond which CVC becomes
detrimental for corporate innovation is never achieved in some firms while other
companies could be very sensitive to splitting the resources between alternative ways of
innovation supply. Capital intensity differs across and within industries; thus for some
corporations establishing a CVC program could pose a greater threat and create more
tensions for the internal R&D that for others. After all, allocating funds to CVC projects
does significantly limit R&D endowment.
The interplay of CVC and R&D is very complex. Engaging in CVC activities
significantly moderates the relationships between R&D intensity and corporate
innovation but this moderation itself is affected by the availability of organizational
slack. In high slack environments CVC intensity positively moderates the relationships
between R&D and corporate innovation: R&D is more effective in advancing corporate
technologies when CVC intensity is higher. In low slack environments the effect of CVC
intensity is the opposite: higher CVC intensity actually hampers the innovative potential
of internal R&D. A company with low level of excess resources may commit a strategic
mistake by attempting to boost its internal innovativeness through introducing a
13
competition to the internal system of innovation supply from CVC-backed ventures:
instead, it may get neither.
Second essay analyzes a broader set of strategic objectives pursued by incumbents
through the CVC programs within the confines of the resource-based view of the firm. In
addition to the rates of corporate innovation it looks at scale efficiency gains that are
likely to follow the increase of demand on the corporation’s products and services from
the new ventures supported by CVC. The essay investigates the effects of the
composition of CVC portfolio of the corporation. It utilizes the classification scheme
offered by Henry Chesbrough (driving, emerging, enabling, and passive investments) and
explores the implications of each of these types of investments for corporate innovation
and scale efficiency gains.
Different types of CVC deals affect the corporate parent in different ways.
Driving investments – i.e., support by the incumbent of new ventures with similar
technologies and high market development potential – are most beneficial for the
corporation. They positively affect both the rates of corporate innovation and scale
efficiency gains. The fact that only one out of six CVC investments targets such ventures
is inconsistent with the espoused ‘strategic’ goals of the majority of CVC programs. This
finding is in line with research on independent venture capitalists investment practices
that demonstrates that espoused criteria are often different from those in use.
Enabling investments are beneficial for the corporation. Both corporate
innovation and scale efficiency stand to gain from investments in companies that provide
strong market development potential for the corporate parent even though they do not
match operational capabilities of the incumbent. That said, positive effect of enabling
14
investments on scale efficiency gains is more than three times higher than the effect of
driving investments. It is thus surprising that enabling investments constitute less than
5% of an average CVC portfolio. While positive effect of enabling investments is
expected in case of scale efficiency gains, it is somewhat surprising for the rates of
corporate innovation. It appears that pull from the market is at least as important to
achieve technical advancement as is push from the technological side. Interestingly,
driving investments which are positively related to corporate innovation are characterized
by strong market development potential.
Emerging investments turn out to be detrimental for the corporation in terms of
both innovation and scale efficiency gains. Apparently, technology push – i.e., surge of
related technologies in the environment that the corporate parent controls – by itself does
not stimulate incumbent’s innovativeness. CVC investments are especially strong
predictors of the technical change when technology push meets market pull: the effect of
driving investments which combine both technology and market fit on the rates of
corporate innovation is more than three times as strong as the effect of enabling
investments that provide market fit alone.
Over half of the investments could be classified as passive: most of the new
ventures supported by CVC fit neither corporate parent’s technologies nor its markets.
Hence, the lack of evidence for the ability of the corporation to add value for their
shareholders in ways that the shareholders cannot do themselves should come at no
surprise: corporations have no superior information about the potential value of most new
ventures they support. These investments are negatively related to both corporate
innovation and scale efficiency gains.
15
Investment risk does not affect scale efficiency gains but hampers the rates of
corporate innovation. Lack of significance for the relationship between R&D intensity
and innovation can be attributed to the effects of temporal lags: as shown in the first
essay, unlike CVC effects that become visible with a two-year temporal lag, R&D effects
are realized quickly but are soon exhausted. Consistent with prior research, this study
demonstrates that organizational slack is important for corporate innovation. At the same
time, it is negatively related to scale efficiency gains. One of the explanations to this fact
may be that organizations without available underutilized capacity may purposefully
target scale efficiency and not technical advancement which requires considerable
resources as a means of improving their factor productivity. It is also of interest that
innovation is associated with larger firms while scale efficiency gains are more of a
smaller companies’ domain. This seems to be consistent with later Schumpeterian views
of innovation being introduced by large established companies; smaller firms are better in
adopting technologies developed by the industry leaders
Third essay analyzes the effect of being a CVC child (having been parented or co-
parented with the help of corporate venture capital) on becoming a CVC parent. It utilizes
organizational imprinting and elements of the resource-based view of the firm as a
theoretical foundation. The study demonstrates that controlling for other things,
corporations with prior history of accepting CVC financing are over 2 times more likely
to initiate CVC programs than corporations without such history, and almost 4 times
more likely to do so if they were parented or co-parented by a corporation with a clear
policy of taking seats on the boards of its CVC children. This could be attributed to the
imprinting in such organizations of corporate venture capital as a legitimate and
16
successful means of accessing know-how developed elsewhere. The imprinting is likely
to happen because at the moment of these firms’ inception both initial environment and
founders – incarnated in the CVC parent – convey through a multitude of channels a
consistent message of corporate venturing being a potent mechanism of tapping into the
knowledge and ideas advanced by technological startups.
Organizational slack and organizational size also turn out to be significant. Larger
corporations with underutilized resources are more likely to initiate corporate venture
capital programs. Industry controls do not demonstrate significance indicating that the
process of becoming a CVC parent varies little across different industries. This
corroborates to some extent the conjecture that organizational imprinting accounts for a
substantial portion of the observed behavior. Namely, corporate venturing is better
described by the dynamics of social and not purely economic or technological processes.
It is imprinting of CVC as a valid instrument in the organization’s toolkit (manifested,
perhaps, in organizational norms, routines, decision criteria, organizational culture) that
makes some organizations and not others to seriously consider initiating CVC programs
in order to strategically screen and support new technologies.
In other words, the study suggests that path dependency applies in the context of
corporate venture capital. Surely, not all CVC parents can trace their heritage to being
CVC children and not all CVC children become CVC parents. At the same time, by
tracing the heritage of incumbent corporations to their founding days the essay
demonstrates that firms started with the help of corporate venture capital are likely to
remain on this path. This sheds light on why some corporations and not others initiate
CVC programs.
17
CVC children are more likely to become CVC parents; they also make better
CVC parents in terms of their ability to absorb the know-how of the new venture which,
all else equal, reflects in higher rates of realized corporate innovation. CVC parents get
more out of the new ventures they support with CVC money if they themselves had prior
experience of accepting CVC financing at founding. Not only do they make better use of
their CVC investments in the immediate perspective, but also their ability to learn from
CVC investments remains superior even when looked at with a two-year forward lag.
At the same time, CVC intensity has a negative effect on the rates of realized
corporate innovation when considered in the immediate perspective. One possible
explanation to this fact is that it takes some time for the corporations to ‘naturalize’
essentially ‘foreign’ innovative ideas possessed by the new ventures. It may also indicate
that organizations need specialized skills to deal with such foreign ideas. In this sense, the
fact that CVC children are superior in their ability to absorb the external know-how is
perfectly in line with this conjecture. This finding seems to indicate that CVC children
may have superior CVC capabilities.
Yet, CVC children are far less likely to effectively utilize their own R&D efforts.
It could be argued that the corporations are forced to make a choice between CVC and
R&D in terms of where to invest their limited resources. At the same time, if this were
also true in the case of CVC children, organizational slack should have been a significant
predictor of the rates of corporate innovation: the tensions would have been less likely to
appear if the organization had significant amount of excess, unutilized or underutilized
capacity. However, organizational slack is not significant in this case. Thus, it appears
reasonable to suggest that it is not the economic substitution of internal source of
18
innovation with the external one that makes CVC children less effective in realizing
innovative potential of their internal R&D efforts. Instead, it could probably be explained
by the imprinted preferences of the CVC children to use new ventures in search of
innovative ideas, preferences that are embodied in their routines, decision making rules
and heuristics, and organizational culture. In this light, it may be interesting to see
whether or not accepting CVC funding at founding impedes future ability of the CVC
child to develop innovative projects internally.
The fact that CVC children are better in realizing the innovative potential of their
CVC investments could also point to their superior CVC capabilities. Since they have
been on the both sides of the CVC process and understand the intricacies of the issues
facing new ventures seeking corporate venture capital support, they are able to provide
adequate technical assistance tailored to the position of the new ventures and address
concerns that the new ventures might have. Accordingly, it is only natural that the time
required for the effects of CVC investments to be reflected in the rates of realized
corporate innovation is shorter for CVC children than for traditionally established
corporations.
Overall, the essays form a cohesive core and help solve some of the challenges
reported by the current CVC literature. The results fit adequately with prior CVC
research, clarify and contextualize important findings reported by the related literature
and shed light on some issues left unanswered by the prior research.
19
References
Anokhin, S. & Schulze, W. 2006. Corporate Venture Capital and Internal R&D: The
Interplay in the Context of Corporate Innovation. Working Paper.
Birkinshaw, J. & Hill, S. 2003. Corporate Venturing Performance: An Investigation Into

the Applicability of Venture Capital Models. Academy of Management Proceedings: B1-
B6.
Burgelman, R. A. 1983. Corporate Entrepreneurship and Strategic Management - Insights

from a Process Study. Management Science, 29(12): 1349-1364.
Chesbrough, H. W. 2002. Making Sense of Corporate Venture Capital. Harvard Business

Review, 80(3): 4-11.
Chesbrough, H. W. & Tucci, C. 2004. Corporate Venture Capital in the Context of

Corporate Innovation. Paper presented at the DRUID Summer Conference 2004 on
Industrial Dynamics, Innovation and Development.
Christopher, A. 2000. Corporate Venture Capital: Moving to the Head of the Class.
Venture Capital Journal, 40(11).
Cohen, W. & Levinthal, D. 1990. Absorptive Capacity: A New Perspective on Learning

and Innovation. Administrative Science Quarterly, 35(1): 128-152.
Dushnitsky, G. 2002. Exploring the Performance Implications of Different Corporate

Venturing Objectives. Working Paper.
Dushnitsky, G. 2004. Limitations to Inter-organizational Knowledge Acquisition: The

Paradox of Corporate Venture Capital. Academy of Management Proceedings: C1-C6.
Dushnitsky, G. & Lenox, M. J. 2005a. When do Incumbents Learn from Entrepreneurial

Ventures? Corporate Venture Capital and Investing Firm Innovation Rates. Research
Policy, 34(5): 615-639.
Dushnitsky, G. & Lenox, M. J. 2005b. When Do Firms Undertake R&D by Investing in

New Ventures? Strategic Management Journal, 26(10): 947-965.
Dushnitsky, G. & Lenox, M. J. forthcoming. Are Firms Profiting from Corporate Venture
Capital? Journal of Business Venturing.
Ernst & Young. 2002. Corporate Venture Capital Report.

http://www.ey.com/global/download.nsf/International/Ernst_&_Young_-
_Global_Venture_Capital_Report_2002/$file/2002_CVC_Report.pdf.
Fleming, G. 2004. Venture Capital Returns in Australia. Venture Capital, 6(1): 23-45.
20
Gompers, P. A. & Lerner, J. 1998. The Determinants of Corporate Venture Capital
Success: Organizational Structure, Incentives, and Complementarities. NBER Working
Paper # 6725.
Gompers, P. A. 2002. Corporations and the Financing of Innovation: The Corporate

Venturing Experience. Federal Reserve Bank of Atlanta Economic Review(Fourth
Quarter): 1-17.
Hellmann, T. 1998. Some Thoughts on the Theory of Corporate Venture Investing,

Corporate Governance Today: 37-43: Columbia Law School.
Hellmann, T. 2002. A Theory of Strategic Venture Investing. Journal of Financial

Economics, 64(2): 285-314.
Keil, T. 2004. Building External Corporate Venturing Capability. Journal of Management

Studies, 41(5): 799-825.
Maula, M. V. J. & Murray, G. 2000. Corporate Venture Capital and the Creation of US
Public Companies: The Impact of Sources of Venture Capital on the Performance of
Portfolio Companies. Paper presented at the 20th Annual International Conference of the
Strategic Management Society.
Maula, M. V. J. 2001. Corporate Venture Capital and the Value-Added for Technology-
Based New Firms. Helsinki University of Technology, Helsinki.
Maula, M. V. J., Autio, E., & Murray, G. 2003. Prerequisites for the Creation of Social
Capital and Subsequent Knowledge Acquisition in Corporate Venture Capital. Venture
Capital, 5(2): 117-134.
Reaume, A. 2003. Is Corporate Venture Capital a Prescription for Success in the

Pharmaceutical Industry? The Journal of Private Equity(Fall): 77-87.
Schumpeter, J. A. 1942. Capitalism, Socialism, and Democracy. New York: Harper and
Brothers.
Wang, K., Wang, C. K., & Lu, Q. 2002. Differences in Performance of Independent and
Finance-Affiliated Venture Capital Firms. The Journal of Financial Research, XXV(1):
59-80.
Zahra, S. A. 1996. Technology strategy and new venture performance: A study of

corporate-sponsored and independent biotechnology ventures. Journal of Business
Venturing, 11(4): 289-321.
Zahra, S. A. & George, G. 1999. Manufacturing Strategy and New Venture Performance:
A Comparison of Independent and Corporate Ventures in the Biotechnology Industry.
The Journal of High Technology Management Research, 10(2): 313-345.
21
APPENDED ESSAYS
22
23
Essay I. Corporate Venture Capital and Internal R&D:
The Interplay in the Context of Corporate Innovation
ABSTRACT
Relationships between corporate venture capital (CVC) and internal research and
development (R&D) have generated much debate in the CVC literature. Some
researchers see them as substitutive while others expect them to be complementary.
Using agency as a theoretical lens this study addresses the controversy and develops new
insights into the nature of the CVC-R&D interplay advancing our understanding of the
mechanisms involved in generating corporate innovation. By considering both economic
and behavioral components of the interplay the framework developed here explains the
contradictory findings reported by the earlier literature.
The analysis demonstrates that the nature of the interplay is complex and shows
the presence of non-linear effects as well as a three-way interaction between CVC, R&D,
and organizational slack in the context of corporate innovation. While both CVC and
R&D appear to be positively related to the rates of corporate innovation, the study
suggests that CVC may have a long-term effect on innovation while R&D is most
influential in the immediate perspective. The results suggest that in the high-slack
environments CVC positively moderates the effect of R&D on innovation while in the
low-slack environments this moderation is negative. The study supports the existence of
the inverted U-shaped relationship between CVC intensity and corporate innovation. The
paper utilizes new methodology to assess realized innovation. Results are compared to
24
other studies that used more conventional innovation measures. By advancing research in
the field of corporate venture capital the study contributes to entrepreneurship, strategy,
and innovation literatures.
25
Corporate Venture Capital and Internal R&D:
The Interplay in the Context of Corporate Innovation
1. Introduction
Corporations are viewed as innovation engines (Schumpeter, 1942); yet
successful innovation often requires access to knowledge that is not available within the
organizational boundaries (Cohen & Levinthal, 1990). To access this knowledge
corporations design sophisticated vehicles that allow identifying and capturing external
innovative ideas. Corporate venture capital (CVC) – equity investments in
entrepreneurial ventures by incumbent firms not made solely for financial gain
(Chesbrough & Tucci, 2004; Dushnitsky & Lenox, 2005a) – appears to be an attractive
organizational arrangement allowing incumbents to tap into the know-how developed
elsewhere (Dushnitsky, 2004): over 450 leading corporations had active CVC programs
within the last decade. Thus, it should be hardly surprising that inquiry into corporate
venture capital seems to have exploded recently (Birkinshaw & Hill, 2003; Chesbrough,
2002; Chesbrough et al., 2004; Dushnitsky, 2004; Dushnitsky & Lenox, 2005b; Ernst &
Young, 2002; Gompers & Lerner, 1998; Hellmann, 1998, 2002; Keil, 2004).
Historically innovation has been a domain of internal research and development
function, or in other words a part of the organizational hierarchy; Zahra (1996) describes
internal product development as a traditional route to corporate entrepreneurship.
Channeling economic activity through hierarchy (e.g., through exclusive reliance on
internal product development), as opposed to market, has its advantages (Coase, 1937;
Williamson, 1985) but comes at costs of which agency-related are perhaps most salient
(Eisenhardt, 1989; Jensen & Meckling, 1976; Zahra, 1996). Adverse selection, moral
26
hazard (shirking) and holdup could become factors to consider given that there is no easy
way to monitor essentially unverifiable research-related efforts of the internal R&D unit
and no way to adequately measure its performance without an external benchmark.
Corporate venture capital – at least in theory – is able to mitigate the prominence of these
problems without undermining the hierarchy advantages in two ways: (1) by introducing
a competition for scarce corporate resources and limiting the availability of the funds for
the internal R&D unit – thus causing it work more efficiently, and (2) by creating a
benchmark in terms of innovation-per-dollar of corporate funds against which the
performance of the internal R&D unit may be assessed. Overall, it challenges internal
R&D unit’s monopoly on producing innovation typical for traditionally structured
corporations with dedicated R&D facilities and by instituting multiple suppliers of
innovative ideas moves the situation into a (more effective and efficient) monopolistic
competition state (Besanko, Dranove, & Shanley, 1996; McConnell & Brue, 2002). Not
only thus corporate venturing serves as a bridge between the corporate core and the
outside world but it also challenges inefficiencies typically associated with extensive
reliance on internal research and development. Zahra (1996) notes that high R&D
spending may simply reflect internal inefficiencies and high agency costs. Depriving
internal R&D from innovation monopoly may help reduce such inefficiencies.
The dominant explanation to why incumbent corporations spend billions of
dollars every year to support new ventures through equity investments is strategic
benefits – in particular, innovation. Whether or not this explanation could stand the
scrutiny of proper statistical testing has not been firmly established in the literature.
While there exists a small body of research that looks at the relationship between
27
corporate venture capital and patenting – which reflects invention or innovative potential
at best – the relationship between equity investment by incumbents into new ventures and
the incumbents’ realized innovation is not well documented. That is, it has been
demonstrated that corporations engaged in corporate venture capital tend to patent more
than their counterparts who do not have dedicated CVC programs (Dushnitsky et al.,
2005a) but whether such patenting reflects true innovation – conceptualized as harnessed
invention (Anokhin, 2004) – or a mere desire to protect new ideas from leaking out is not
clear. It has been argued that not all patents are acted upon; not all inventions are
patented (Carlsson & Fridh, 2002). Some ideas are patented just to keep competitors from
pursuing the respective opportunities, whereas some ideas are pursued in secrecy with
patent application never filed for. For this reason, although patenting and innovation are
related, this relationship is far from perfect. This paper employs a “cleaner” measure of
innovation conceptually close to the views of Edith Penrose (1959) and her attention to
resource allocation efficiency. The measure utilizes a technique of the factor productivity
change decomposition originally developed by Malmquist (1953), adapted to productivity
measures by Caves and his colleagues (Caves, Christensen, & Diewert, 1982), refined by
Fare and his collaborators for economics applications (Fare, Grosskopf, Norris, & Zhang,
1994) and recently adopted by strategic management (Delmas & Tokat, 2005; Durand &
Vargas, 2003) and innovation (Thursby & Thursby, 2002) literatures to assess overall
resource allocation efficiency and estimate innovation (technical advancement).
Despite multiple calls throughout the last decade (Gompers, 2002), the nature of
the interplay between internal R&D and CVC both of which are expected to affect
innovation has not been fully attended to by the researchers. Extant empirical research
28
fails to describe let alone explain the directionality of such interplay and its implications
for corporate innovation. There exists almost no literature that looks at this question
specifically or goes beyond conventional thinking and tests the relationship statistically
(see Chesbrough et al. (2004) for an exception). While it is commonplace to include
internal R&D as a covariate into models assessing the impact of externally-oriented
corporate venturing (Dushnitsky et al., 2005a), how one affects the other is still largely
unknown. Several researchers have hinted at the existence of such relationship to date,
and provided mostly (well-grounded) speculations as to how the interaction between the
two may look (Chesbrough & Socolof, 2000; Gompers et al., 1998). Unfortunately, these
speculations are not in agreement: while some suggest that the corporation has to
explicitly choose between external and internal research (Hellmann, 2002), others
maintain that the relationships are rather complementary (Chesbrough et al., 2004).
Using the research by Gompers and Lerner (1998), Dushnitsky and Lenox
(2005b; 2005a), Chesbrough and his colleagues (Chesbrough, 2000, 2002; Chesbrough et
al., 2004), Keil, Zahra and Maula (2004) and others as a starting point, this paper borrows
from economics and innovation literatures to shift the accent from potential to realized
(harnessed) innovation and attempts to settle the controversy with regards to the interplay
between corporate venture capital and internal research and development providing a
significant advancement in theoretical development. It supplements the theoretical
advances with rigorous empirical testing. Section 2 provides a literature overview and
develops the hypotheses. Methodological issues are covered in Section 3. Results are
presented in Section 4 and discussed in Section 5. The paper concludes with the
discussion of limitations of the current study and suggestions for future research.
29
2. Literature Review and Hypotheses Development
Although many factors other than R&D expenditures may be responsible for
performance differences – such as the firm’s unique capability to deploy or transform its
resources that result in sustainable competitive advantage (Dierickx & Cool, 1989;
Leonard-Barton, 1995; Yeoh & Roth, 1999) – commitment to R&D spending is
necessary for innovation (Capon, Farley, & Lehmann, 1992; Hambrick & MacMillan,
1985; Maidique & Hayes, 1984). Innovative output is positively related to R&D (Acs &
Audretsch, 1988). In research practice, absent of direct measures of innovation, research
and development-related measures sometimes are even used as a proxy for innovation
itself or otherwise equated with innovation (Cohen, Levin, & Mowery, 1987; Sah &
Stiglitz, 1987).
R&D intensity has also been shown to positively affect the rates of corporate
patenting (Dushnitsky et al., 2005a); the relationship is particularly strong for high-
quality, influential patents that receive a large number of citations. In the present study,
however, we are interested in the realized innovation, and not just knowledge codification
represented by patenting. After all, corporations invest their money not to patent more but
to bring about technical advancement – that is, the ability to achieve more with a
comparable resource bundle or use less resources to produce the same amount of output.
Traditional proxies for innovation – such as patents, patent citations or even new products
launch – cannot fully capture the notion of realized innovation. Innovation does not
necessarily include launching new products. The introduction of a new good (or of a new
quality of a good) is only one of five possible types of innovation. Others include the
30
introduction of a new method of a production, the opening of a new market, the conquest
of a new source of supply of raw materials or half-manufactured goods, and the carrying
out the new organizations of a industry (Schumpeter, 1934). Patents, in turn, are a
somewhat noisy measure of innovation as not all inventions are patented and not all
patents are acted upon(Carlsson et al., 2002). While macro-level research has
demonstrated that R&D is positively and statistically significantly related to realized
innovation (Kim & Lee, 2004) it remains to be seen whether the relationship holds at the
micro-level as well. Building on the classic innovation management literature (Capon et
al., 1992; Hambrick et al., 1985; Maidique et al., 1984) and newer macro-level studies
(Kim et al., 2004) we hypothesize the following relationship:
Hypothesis 1. Internal R&D intensity is positively associated with the rates of corporate
innovation.
While information may be “sticky” (von Hippel, 1994), very rarely could it be
kept completely intact by and insulated from the outer world. This fact is best attested by
the very existence of the institution of patents. Despite varying degree of environmental
tenacity – or the strength of industry appropriability regime (Levin, Klevorick, Nelson, &
Winter, 1987; Shane, 2001) – sooner or later knowledge tends to diffuse. In case of
corporate venture capital investment it is likely to happen sooner rather than later as most
corporations actively facilitate the spillover and seek opportunities to absorb and
integrate the supported ventures’ know-how (Dushnitsky, 2004; Keil et al., 2004). For
instance, corporate parents almost always take seats on their investees’ boards and thus
31
could observe and learn from the startups (Anokhin & Schulze, 2006). This is particularly
true when funded ventures fit the corporate profile in one way or another. Dushnitsky et
al. (2005a) demonstrated that corporate venture capital intensity positively affects
corporate innovation potential (proxied with patenting) when corporation explicitly
pursued strategic and not financial goals. Similarly, one could expect CVC intensity to
reflect in the rates of realized corporate innovation:
Hypothesis 2. CVC intensity is positively associated with the rates of corporate
innovation.
Innovative ideas can be produced in-house (internal R&D) or accessed outside
(e.g., through corporate venture capital program); organizations do not have to limit
themselves to one alternative only. Despite prevalence of the in-house organization of
corporate innovative activities throughout economic history, providing internal R&D unit
with a de-facto monopoly on supplying innovation has its drawbacks. As classic
economics-based industrial organization literature suggests, monopolies never act
efficiently and ultimately pass the inefficiency costs down to the consumers (Besanko et
al., 1996). The idea has also been acknowledged by the corporate entrepreneurship
literature: Zahra (1996) argued that high R&D spending may simply indicate (1) internal
inefficiencies and (2) high agency costs rather than aggressive risk-taking attitude or
successful innovative activities.
Building on the first part of Zahra’s argument and developing further the
industrial organization analogy, this study suggests that to ensure efficiency the
32
corporation needs to destroy the R&D unit monopoly. As economics posits, having (even
a limited number of) competing players can substantially improve efficiency as compared
to the state of monopoly (McConnell et al., 2002). In the context of corporate innovative
activities structuring this goal may be achieved through (1) establishing multiple
corporate R&D centers or (2) supporting multiple technological initiatives via means like
corporate venture capital. Economics warns of potential collusion pitfalls when having a
very limited number of players; ensuring geographical dispersion of corporate R&D
centers could address this issue at least to some extent. Indeed, as Tripsas (1997) reports,
having multiple geographically distributed research sites spurs corporate innovation,
helps to overcome organizational inertia, provides a source of variation, and thus
facilitates the development of new technological capabilities and technical advancement.
Second option (multiple technological initiatives supported through CVC programs) has
also been implemented by many leading corporations of the world in the last forty years.
Introducing competition into innovation producing processes, according to classic
economics, is a way to rid innovative activities of inefficiencies.
The second part of Zahra’s argument suggests that internal research and
development may be associated with high agency costs. Agency theory focuses on
inefficiencies arising from the contracting relations between one party (the principal)
engaging another party (the agent) to perform certain tasks on its behalf, which involves
delegating some decision making authority to the agent (Eisenhardt, 1989; Jensen et al.,
1976). The idea is that both parties of the contract are self-interested, rational,
opportunistic, and risk-averse. Because of that, the agent’s (opportunistic) behavior is not
necessarily consistent with the principal’s best interests (Eisenhardt, 1989). Opportunistic
33
behavior of the agent manifests itself through adverse selection (e.g., agent’s
misrepresentation of its capabilities), moral hazard (e.g., agent’s shirking due to the
unobservability of its actions by the principal) and hold-up (e.g., renegotiating contract
terms by one party after another party’s investments into relation-specific assets due to
incomplete contracts and inability to foresee all contingencies ex ante) (Aghion & Tirole,
1994). In order to address the opportunistic behavior of the agent the principal needs to
incur monitoring costs (monitoring the performance of the agent). In turn, the agent
incurs bonding costs (i.e., the cost of the guarantee it provides to the principal that it
would not engage in opportunistic behavior). Finally, any misalignment left between the
agent’s and the principal’s interests implies the residual loss on the part of the principal
(i.e., the principal’s value is not maximized) (Eisenhardt, 1989; Jensen et al., 1976).
The typical setting where agency problem applies is the relations between owners
and managers (Eisenhardt, 1989). However, it could easily be extended to apply to the
relations of the corporation top management and the corporate R&D unit: internal R&D
can be conceptualized as an agent of corporate management (principal) in charge of
technical advancement. Here, adverse selection may become an issue if the R&D
department starts a project beyond its area of expertise. Moral hazard is a typical problem
whenever the agent’s actions are not verifiable, which is often the case with highly
complicated research and development projects; for this reason rewarding the observable
outcomes rather than non-verifiable behavior may be a solution (Eisenhardt, 1985). Hold-
up problems occur when apparently unsuccessful internal projects are not deprived of the
corporate funding even if no useful results are achieved despite significant corporate
resource spending.
34
It appears that establishing a corporate venture capital program may assist in
addressing at least some of the concerns associated with the agency problem. In essence,
CVC programs challenge internal R&D units’ monopoly on producing innovation. In his
study of Lucent’s Bell Labs, Chesbrough (2000) provides anecdotal evidence to the
effects of such challenge: “The …[CVC] also appears to be achieving its goal of serving
as an impetus for Bell Labs technologies to move off of the shelf… [CVC’s] early
interest in technologies has caused some Bell Labs technologies to move directly into the
business groups that might otherwise have been overlooked by those businesses” (p. 44-
45). While encouraging corporate venturing may in principle be dangerous for the threat
of opportunism (Burgelman, 1983), limiting innovation entirely to the corporate R&D
may be substantially more problematic. Negativities typically associated with monopoly
are only one part of the problem. Without requisite diversity one of the intrapreneurship
major goals – strategic renewal – may never be achieved (Burgelman, 1983).
Establishing a corporate venture capital programs is particularly helpful in
reducing adverse selection and moral hazard (shirking). Allocation of corporate funds to
corporate venturing threatens the availability of funds to the internal R&D unit and thus
spurs competition in line with economics reasoning outlined above. In other words, CVC
programs could be seen as substitutive rather than complementary to the corporate R&D
programs; introduction and expansion of funds allocation to the CVC programs is likely
to adversely affect the size of corporate R&D budgets. Therefore:
Hypothesis 3. CVC intensity is negatively related to internal R&D intensity.
35
When the incumbent corporation makes a strategic investment in a particular area
it effectively commits itself not to make an identical (or similar) rival internal investment
(Hellmann, 2002). Accordingly, to warrant allocation of sufficient corporate funds,
internal R&D unit has to be particularly selective with regards to the choice of projects to
develop dependent on the level of corporate venture investment activity. Thus, presence
of the CVC alternative to a purely hierarchical supply of innovation is likely to reduce the
strength of the adverse selection problem. The result is increased R&D unit’s
productivity and improved quality of the research output. We define it as the “economic
component” of the CVC-R&D interplay. Second, equity investments in new
technological startups create a benchmark in terms of ‘innovation-per-dollar-of-
corporate-funds’ against which internal R&D unit performance may be evaluated. This
helps reduce (to an extent) the agent’s shirking intentions creating a powerful incentive
for the R&D unit to act more efficiently. In other words, even if engaging in CVC
programs does not reduce R&D financial support, the CVC-R&D interplay still has its
“behavioral effect” on the R&D unit management and leads to desirable outcomes.
However, both these effects may be potent only if the organization has sufficient
amount of organizational slack, or a cushion of excess resources that can be used in a
discretionary manner (Bourgeois, 1981). The reason is that establishing a CVC program
by the incumbent corporation creates tensions for its in-house R&D group including
competition for corporate resources and innovation leadership. To release tensions,
organizational slack may be instrumental. This is true both in general, and in the CVC
context as well. Chesbrough et al. (2004) noted that organizational slack may be an
important determinant of the direction of the CVC-R&D interplay: potentially slack may
36
determine both CVC and R&D spending. When firms have extra resources, they might
spend more on R&D and other things, such as CVC. When firms have no extra resources,
they may cut back on both. In other words, slack may affect the levels of both CVC and
R&D. Besides, higher slack availability creates an environment conducive for innovation
while low slack, on the contrary, may cause firms to withdraw from potentially promising
innovative projects. Slack facilitates innovation regardless of the hierarchical
arrangements that innovative activities may be subjected to (Anokhin, 2004; Cheng et al.,
1997; Lawson, 2001). CVC context is no exception. Namely, when organizational slack
is high we expect R&D to increase its productivity as the levels of CVC intensity go up.
At the same time, we suggest that when the organization has no extra resources by
investing in outside projects such as CVC it may negatively affect the productivity of its
internal R&D. Stated formally:
Hypothesis 4a. CVC intensity positively moderates the effect of internal R&D intensity on
the rates of corporate innovation when organizational slack is high.
Hypothesis 4b. CVC intensity negatively moderates the effect of internal R&D intensity
on the rates of corporate innovation when organizational slack is low.
Overall, we propose that the pattern of the CVC-R&D interplay is complex. CVC
activity in addition to its direct positive effect on corporate innovation interacts with
internal R&D function in a complicated way: negatively through reduction of the funds
available to the R&D unit, and positively (or negatively) through the moderation of
37
R&D-Innovation relationship when slack is high (low). Thus, the relationship between
CVC and innovation is subject to two types of forces pulling it in the opposite directions.
We expect this to result in existence of the optimal size of CVC portfolio whereafter
further increase in the number of ventures supported through CVC does not lead to more
innovation due to confounding the effects of R&D and may actually hamper innovation.
In other words, we suggest that up to a certain threshold CVC activity is beneficial for
corporate innovation: due to knowledge spillovers and economic and behavioral
components of the CVC-R&D interplay corporate innovativeness stands to gain from
CVC investments. Past the threshold, however, pure effect of CVC on corporate
innovation is rather negative due to exhaustion of funds otherwise available to internal
research and development. Therefore:
Hypothesis 5. There exists an inverted U-shape relationship between CVC intensity and
the rates of corporate innovation.
It should be noted that allocation of corporate funds to external technological
initiatives is potentially plagued with agency problems of its own kind; however, in
corporate venture capital practice monitoring costs are typically spread out between
several participants of the investment syndicate and the hold-up problem is minimized
due to investment-round-based approach to funds allocation.
38
3. Data and Method
3.1. Data
It has been noted that the corporate venture capital cycle historically runs about
every ten years (Block & MacMillan, 1995; Chesbrough et al., 2004). To date, there have
been three (very different in kind and magnitude) waves of captive venture capital
investments. First (mid-1960s through early 1970s) was driven by the corporations that
expected to utilize their industry expertise to differentiate successful from unsuccessful
ventures and obtain financial returns similar to those of independent venture capital
funds. Second wave (late 1970s through late 1980s) was driven by the finance-affiliated
institutions which for the first time were allowed to invest in high-risk assets including
venture capital. Third wave (mid-1990s to early 2000s), similarly to the first one, was led
again by the corporations; only this time the main purpose was strategic and not financial
benefits (Gompers, 2002). Currently, there are certain indications that a fourth CVC cycle
is about to start (Red Herring, 2005). Given the qualitative differences between the
waves, to reduce information noise one needs to focus the empirical investigation on a
single wave. Of three waves only the last one was clearly strategically oriented and is
thus most suitable for the study of the relationships between CVC and corporate
innovation. Thus, in our investigation we analyze the period of 1998-2001 which covers
growth, peak, and decline of the third wave.
We use multiple secondary sources to test the hypotheses1. VentureXpert database
by Venture Economics is by far the most popular source of data in the corporate venture
capital research. It was employed by Gompers et al. (1998), Dushnitsky and his
1
Information on CVC used in this paper is borrowed from Zahra, Schulze, & Anokhin (2005)
39
colleagues (Dushnitsky, 2004; Dushnitsky et al., 2005b, 2005a; Dushnitsky & Lenox,
forthcoming), Keil et al. (2004), Maula & Murray (2000) and others. This database
contains a comprehensive coverage of investment, exit, and performance activity in the
private equity industry form 1969 to the present time2. Despite its popularity, there are
certain challenges in using the VentureXpert data. Lerner (1994) claims that it has
significant bias: single venture investment rounds, particularly in more mature firms, are
often recorded as several observations. According to his calculations, the database reports
28 percent more rounds than actually occurred. The problems are likely to be more severe
in later rounds, which typically have more investors (Lerner, 1995).
We employ two mechanisms to deal with this issue. First, we aggregate the
VentureXpert data at the annual level. This way, even if a particular single round has
been recorded twice or more due to funding not being disbursed at once, the annual
aggregation should provide a more accurate account. Second, we use an alternative data
source – multiple editions of Corporate Venturing Directory & Yearbook by
AssetAlternatives (Barry, 2000, 2001, 2002) (hereafter “the Yearbook”) – to cross-check
the investments reported by VentureXpert and their properties. The Yearbook has been
used in corporate venture capital research before as well (Birkinshaw et al., 2003;
Chesbrough et al., 2004; Dushnitsky, 2002; Gompers, 2002). It also may double-count
2
Following CVC literature we used the following VentureXpert categories to define the population of
firms engaged in corporate venture capital activities: Non-Financial Corp Affiliate or Subsidiary
Partnership, Venture/PE Subsidiary of Non-Financial Corp, Venture/PE Subsidiary of Other Companies
NEC, Venture/PE Subsidiary of Service Providers, Direct Investor/Non-Financial Corp, Direct
Investor/Service Provider, SBIC Affiliate with Non-Financial Corp, and Non-Financial Corp Affiliate or
Subsidiary
40
certain deals on rare occasions (e.g. a single investment by Mitsui into Radix Wireless
Inc. also known as BeamReach Networks Inc. in April 2000 has been recorded as two
separate events) so we carefully check for and correct such occurrences.
Finally, we merge our database with the annual firm-level accounting and
financial data from Standard & Poor’s Compustat. Since the data reported in Compustat
relate to a financial and not a calendar year of the corporation, we do not use annual
aggregates reported by VentureXpert directly but rather look at the exact dates of
particular deals to match them to appropriate financial years. Thus, for a corporation with
financial year starting in March and ending in February we would consider a CVC
investment made in January of 2000 as a part of year 1999. The merger of VentureXpert,
the Yearbook, and Compustat is based on companies’ tickers and descriptions and yields
a sample of 163 corporations that engaged in corporate venture capital investments
during years 1998 to 2001.
3.2. Variables
3.2.1. Dependent variable
Rates of Corporate Innovation (RCI). Corporate innovation measure is derived
from the nonparametric programming technique known as Data Envelopment Analysis
(DEA) adopted by Fare and his colleagues (Fare et al., 1994) to examine productivity
growth. This technique long used in economics has recently made its way into strategic
management (Delmas et al., 2005; Durand et al., 2003) and innovation (Thursby et al.,
2002) literatures. The idea behind the technique is as follows. Companies use different
combinations of homogenous inputs (e.g., labor and capital) to produce similar outputs
41
(represented by sales). Some firms use less labor than others, some use less capital, while
yet others use relatively large quantities of both labor and capital to produce the same
amount of output. By comparing all observable combinations of inputs and outputs DEA
calculates the ‘best practice’ production frontier which reflects the best known
technology of the time. Corporations which are efficient would find themselves on the
frontier; inefficient companies, in turn, would be at a distance from the frontier, and this
distance is conceptualized as a measure of their inefficiency. There is no assumption on
the underlying productive function (technology) and no constant return to scale
constraints in DEA. Over time, as the companies improve technologies they employ they
require less labor and capital to produce the same amount of output. Figure 1 provides an
illustration of the intertemporal production frontier in the computer and electronic
product manufacturing industry in 1998-2003: one can easily see the movement of the
corporations from the upper-right corner (large quantities of both labor and capital
required to produce a unit of sales) to the lower-left corner (smaller quantities of both
resources). All numbers used to create Figure 1 are adjusted for inflation based on annual
CPI indices reported by the Bureau of Labor Statistics.
--------------------------- Insert Figure 1 about here ---------------------------
By tracking shifts of the frontier from year to year, and movement of the
companies with respect to the frontier, it is possible to decompose the corporation’s total
factor productivity change into two components – technical change and efficiency
change. The algorithm was first offered by Malmquist (1953) and is thus known as
42
Malmquist decomposition. The former component (technical change) is generally
accepted as a measure of innovation (Fare et al., 1994; Thursby et al., 2002). It is superior
to patenting as not all patents are acted upon whereas technical change in the production
frontier context always reflects the rates of realized innovation. Basically, it reflects
improvements in resource allocation efficiency due to technical advancement regardless
of the nature of advancement (new process, new materials, etc.) and thus is in line with
the entrepreneurship literature (Penrose, 1959; Schumpeter, 1934). Appendix 1 provides
a brief description of and further references on DEA and Malmquist decomposition.
Following the seminal American Economic Review article by Fare and his
colleagues (Fare et al., 1994), we use physical assets and labor as inputs, and sales as an
output. This approach is commonplace in the DEA literature. Data for estimating the
frontiers and performing Malmquist decomposition is directly available from Compustat.
As industries are likely to differ with respect to their innovative activities patterns, we
identify twenty subindustries based on NAICS classification (at 2-6 digits aggregation)
and calculate the rates of technical advancement separately in each of them. All active
firms operating in the subindustries of interest without missing data as reported by
Compustat (12,816 firm-year observations overall) are analyzed at that step to ensure
most accurate results. Appendix 2 presents the composition of the subindustries used in
this study. As it may take time for innovation to be fully harnessed, the models are tested
with different temporal lags (immediate perspective, one- and two-year forward lags).
We scale the dependent variable in such a way that values over 100 indicate adoption of
superior technologies as compared to the previous year.
43
3.2.2. Independent variables
Internal R&D intensity is defined as the ratio of R&D outlays to the dollar amount
of the corporation’s assets. Adjustment for the industry average is made for each of the
twenty subindustries. The data is directly available from Compustat. It has been
explained that there is no relation between the size of CVC investment which often
simply depends on the investment round and strategic benefits to the corporation
(Reaume, 2003). For that reason we use the number of the new ventures supported each
year by the incumbent corporation rather than the dollar amount invested as our measure
of CVC Intensity. It has also been shown to be highly correlated with the dollar amount:
Anokhin et al. (2006) report the correlation coefficient of 0.85. Following Singh (1986),
we operationalize organizational slack as current ratio or the ability of the firm to meet
current obligations (current assets divided by current liabilities). The data is directly
available from Compustat.
3.2.3. Control variables
Risk is an important determinant of venture capital activity outcomes. As
compared to independent venture capital, corporate venture capital is claimed to invest in
more established, less risky companies (Ernst & Young, 2002; Riyanto &
Schwienbacher, 2005). In line with the venture capital literature, we use the corporation’s
preferred investment round as a measure of risk (Ernst & Young, 2002; Fredriksen &
Klofsten, 2001). We dichotomize investments into relatively risky (seed and early stages)
and relatively non-risky (extension, later, and balanced stages). To create the variable we
use the data from both the Yearbook and VentureXpert. To account for the temporal
44
effect we create a dummy variable for the year 2000 which was characterized by the
dot.com market correction. There are also significant between-industry differences:
according to European Venture Capital Journal estimates more than 70 % of all corporate
venture capital investments are allocated to computer-related industries,
telecommunication hardware and internet technology (EVCJ, 2002). Level of CVC
activity in the industry creates pressure to innovate (Zajac, Golden, & Shortell, 1991).
Besides, environments in which businesses operate might have other ways of affecting
the relationships between the variables of interests. For this reason we create industry
dummies and aggregate the subindustries into six industrial groups. Finally, we control
for the firm size with the log of sales.
3.3. Method
The dataset covers multiple years, thus panel data techniques should be used.
Malmquist estimates of realized corporate innovation may violate the independence
assumption required for most OLS-based techniques (Thursby et al., 2002). To account
for the non-independence we employ feasible generalized least square estimator (Greene,
2003). Following Beck & Katz (1995) we take a conservative stand and model first-order
temporal correlation for our models. It has been argued that FGLS estimators best work
in temporally dominated models with the number of time points substantially exceeding
the number of units; otherwise confidence in the findings may be falsely inflated (Beck et
al., 1995). We address this concern by following a procedure suggested by Wiggins
(2001) and re-estimate our final model with the Huber-White estimates of variance
(standard errors). Remarkably similar results are obtained.
45
4. Results
In Table 1 we present descriptive statistics and the correlation matrix for the
measures used in this study. Overall, the magnitude of correlation coefficients is similar
to those reported by Chesbrough et al. (2004). On average, corporations in our sample do
experience technical advancement (mean RCI=107.51) and support a little over six
different ventures per year through CVC although some of the most active companies
may invest in over 250 new ventures annually (e.g., Intel in year 2000). About 16 percent
of the corporations engaged in CVC practices tend to support riskier projects. The
majority (84%) concentrate on less risky expansion, later, and balanced stages.
Interestingly, this study reproduces the (unexpected) negative association between
organizational slack and R&D intensity reported by Chesbrough et al. (2004). Following
these authors, we suggest that the possible explanation lies in that firms that cut back on
their R&D spending have more cash in the current year.
--------------------------- Insert Table 1 about here ---------------------------
Results of hypotheses testing are presented in Table 2 which is organized as
follows. Model 1 is a test of the effect CVC intensity has on R&D intensity. Models 2-7
estimate relationships between CVC intensity, R&D intensity, and the rates of corporate
innovation (RCI). Model 2 considers RCI in the immediate perspective, Models 3-7 –
with a two-year forward lag. Models 4-7 build on Model 3 by adding sets of controls
(industry and year) and a square term for CVC intensity. Model 8 adds to the set of
46
predictors a three-way interaction between CVC intensity, R&D intensity, and
Organizational Slack. Model 9 re-estimates Model 8 with the Huber-White estimates of
variance (standard errors) to address to possibility of falsely inflating the confidence in
the findings reported by Model 8. It utilizes first differencing technique to account for
firm-specific effects (Wooldridge, 1999). Overall, Hypotheses 1-2 are tested in Models 2-
9, Hypotheses 3 – in Model 1, Hypotheses 4a and 4b – in Models 8-9, and Hypothesis 5 –
in Models 5-9.
Hypothesis 1 finds support in Models 2, 6, and 7. Hypothesis 2 is supported
throughout Models 3-9. Only in the immediate perspective (Model 2) is the effect of
CVC intensity on the technical change (rates of corporate innovation) not positive
significant. Hypothesis 3 is supported by Model 1: CVC intensity is negatively related to
R&D intensity. Hypotheses 4a and 4b are supported by both Models 8 and 9. The three-
way interaction hypothesis is supported. To ease the interpretation of the interaction we
plot it in Figure 2. Finally, Hypothesis 5 is supported by Models 5, 6, 8, and 9. All
Models demonstrate good fit as suggested by highly significant (p<0.001) test statistics
with exception of Model 3 which is significant at p<0.05 level.
47
5. Discussion
Overall, the results lend support to the conjecture that R&D intensity is positively
related to realized corporate innovation as measured by the rates of technical change.
This agrees to the results reported by the earlier studies: indeed, commitment to R&D is
necessary for innovation. Upon examining the regression coefficients and their respective
significance levels it appears that the effect of R&D intensity on the rates of corporate
innovation may be more prominent in the immediate (Model 2) and not in the long-term
perspective. This is in stark contrast to the effect of CVC intensity which is a particularly
significant predictor of the rates of corporate innovation when considered with a forward
lag of one to two years. This could indicate that firms adopt the ideas generated by their
own R&D function faster than those produced by the outside ventures. If true, it may
suggest the necessity of developing additional vehicles and building absorptive capacity
to accelerate adoption of the externally accessed know-how.
This longer timeframe required for the effects of CVC to be realized by the
corporate parent may also necessitate the need for developing integrative capabilities. To
be able to internalize the know-how developed elsewhere, the firm must be able to
properly capture the external idea and to effectively combine it with the knowledge that
exists in-house. Among other things, that may require considering special governance
arrangements for the CVC unit, careful composition of the CVC unit in terms of the
functions represented in the CVC team, developing policies with respect to taking seats
on the boards of the new ventures supported with the corporate venture capital, and so on.
The question of leadership in terms of the organizational hierarchy is also important:
48
having an overly senior or an overly junior champion of the CVC unit may turn
detrimental for the success of the project.
Alternatively, much shorter time required for the effects of internal R&D to be
realized by the corporation could also indicate that internally advanced ideas are simply
focused on the short term as they need to address problems that exist here and now. If
that is true, it may suggest that internal research and development could develop
innovation myopia of some sort. Applied research without the fundamental component is
unlikely to sustain itself for long. In this case, if the focus of corporate R&D has indeed
shifted to the short-term perspective, the corporations may consider developing strong
relationships with the institutions that engage in fundamental research such as
universities and research labs. This should also help build external integrative capability
and thus positively affect the adoption by the incumbent corporations of the ideas
developed elsewhere.
More importantly, however, this finding suggests that qualitatively R&D and
CVC are not perfect substitutes. To ensure successful technical advancement in short-
and long-term perspectives corporations may be better off supporting both own and
outside suppliers of innovative ideas. Thus, the very question of ‘complementary’ vs.
‘substitutive’ relationships between R&D and CVC in the context of corporate innovation
may receive a whole new meaning. The question may change from ‘whether to establish
a CVC program’ to ‘how much funds dedicate to the program’ and ‘what types of
ventures to support’. CVC investments are not homogenous; yet, little is known about the
effect of each of the different types of investments on corporate innovativeness.
49
Positive effects of CVC intensity on realized corporate innovation are supported
throughout Models 3-9. This corroborates Dushnitsky et al.’s (2005a) recent findings and
extends the implications of their work beyond the realm of potential into realized
innovation. This study, however, goes further and examines the presence of quadratic
effects in the relationships between CVC intensity and innovation. As hypothesized, there
exists an inverted U-shaped relationship between the level of CVC intensity and realized
corporate innovation. Together with the findings reported above it suggests that it is
possible to determine the optimal size of CVC portfolio and to balance CVC and R&D
activity of the firm in such a way as to maximize corporate innovation. The question
remains whether or not this effect is subject to some inter-firm differences that we do not
capture. It may be that the threshold beyond which CVC becomes detrimental for
corporate innovation is never achieved in some firms while other companies could be
very sensitive to splitting the resources between alternative ways of innovation supply.
Capital intensity differs across and within industries; thus for some corporations
establishing a CVC program could pose a greater threat and create more tensions for the
internal R&D that for others. After all, our analysis demonstrates that allocating funds to
CVC projects does significantly limit R&D endowment.
The interplay of CVC and R&D is indeed very complex. Engaging in CVC
activities significantly moderates the relationships between R&D intensity and corporate
innovation but this moderation itself is affected by the availability of organizational
slack. In high slack environments CVC intensity positively moderates the relationships
between R&D and corporate innovation: R&D is more effective in advancing corporate
technologies when CVC intensity is higher. In low slack environments the effect of CVC
50
intensity is the opposite: higher CVC intensity actually hampers the innovative potential
of internal R&D. A company with low level of excess resources may commit a strategic
mistake by attempting to boost its internal innovativeness through introducing a
competition to the internal system of innovation supply from CVC-backed ventures:
instead, it may get neither.
6. Limitations and Conclusions
Our research design relies on secondary longitudinal data collected from multiple
sources. Thus, some of the concerns such as common method bias typically associated
with primary cross-sectional data collection are less of a problem here. We control for
confounding effects due to cross-industry differences, intra-industry factors (e.g.,
temporal changes), and intra-industry heterogeneity in firms’ attributes. Yet, there remain
issues that potentially could threaten the validity of our hypotheses testing. First, there
might be different sampling concerns. Since we have no control over the initial data
collection process we cannot be sure that all assumptions are met, and the CVC programs
studied are representative of the overall population. Lerner (1994; 1995) indeed reports
certain problems with the VentureXpert database. While we try to address this concern by
using alternative data sources and conservatively cross-checking the entries in
VentureXpert and the Yearbook, these alternative sources as we show are also not
entirely free from misreporting and omissions.
The nature of the dependent variable and its newness in the field of (corporate)
entrepreneurship and strategy is also a factor to consider. More studies subscribing to the
same framework are needed to generate the critical mass of knowledge about this
51
measure of realized innovation. At the same time, the benefits of shifting attention from
invention to innovation are numerous. Patents and patent-related proxies for innovation
are neither comprehensive (three out of five types of innovation are hardly patentable)
nor always appropriate (small and young firms often do not have patents; thus this
measure of innovation is biased towards larger and older firms). Considering real-life
evidence of the corporations’ activities, and being able to compare the ways in which the
firms combine resources to produce their outputs enables us to estimate the rates of
realized innovation and thus adds to the strategic management toolkit. Particular value of
this approach is in that it helps to empirically distinguish between the notions of
invention and innovation, which is harnessed invention. After all, corporations do not
invest money in research and development and in outside ventures to patent more; what
they look for is the ability to increase effectiveness of their resources deployment. The
DEA-based technique used in this paper provides a means through which realized and not
potential innovation may be tracked.
Remaining concerns have to do with possible model misspecification due to
omitted variable bias. Unless uncorrelated with other independent variables, omitted
variables may cause distortion of some estimates reported here. Of particular interest are
alternative modes of sourcing technological know-how such as licensing agreements and
alliances of different sorts. There is currently some work being done in this area by other
researchers that explores these issues in more details; to the best of our knowledge our
results should hold.
It is also important to look at the types of particular CVC deals as not all deals
have equally strong connection to the firm’s operational capabilities and pursued markets.
52
It is likely that different types of investments have different implications for corporate
innovation. The same holds true for the notion of deals syndication. This particular
property of VC practices may have important implications for the rates of the firm’s
technical advancement. Accounting for the deals heterogeneity and syndication practices
are thus necessary steps in further untangling of the effects of corporate venture capital
on corporate innovation. We leave these steps to the future research.
53
References
Acs, Z. J. & Audretsch, D. B. 1988. Innovation in Large and Small Firms: An Empirical
Analysis. The American Economic Review, 78(4): 678-690.
Aghion, P. & Tirole, J. 1994. The Management of Innovation. Quarterly Journal of

Economics(November): 1185-1209.
Anokhin, S. 2004. Excess Technology Commercialization: Making Distant Knowledge

Work. Paper presented at the Academy of Management Meeting, New Orleans, LA.
Anokhin, S. & Schulze, W. 2006. Do CVC Children Make Better CVC Parents? Working
Paper.
Barry, D. G. (Ed.). 2000. The Corporate Venturing Directory and Yearbook. Wellesley,
MA: Asset Alternatives.
Beck, N. & Katz, J. N. 1995. What to Do (and Not to Do) with Time-Series Cross-
Section Data. American Political Science Review, 89(634-647).
Besanko, D., Dranove, D., & Shanley, M. 1996. The Economics of Strategy. New York:
John Wiley & Sons, Inc.

B6.
Block, Z. & MacMillan, I. 1995. Corporate Venturing: Creating New Businesses withing
the Firm. Boston, MA: Harvard Business School Press.
Bourgeois, L. J., III. 1981. On the Measurement of Organizational Slack. Academy of

Management Review, 6(1): 29-39.

Capon, N., Farley, J. U., & Lehmann, D. R. 1992. Profiles of Product Innovators Among
Large U.S. Manufacturers. Management Science, 38(2): 157-169.
Carlsson, B. & Fridh, A. 2002. Technology Transfer in United States Universities: A

Survey and Statistical Analysis. Journal of Evolutionary Economics(12): 199-232.
54
Caves, D. W., Christensen, L. R., & Diewert, W. E. 1982. The economic theory of index
numbers and the measurement of input, output, and productivity. Econometrica, 50:
1393-1414.
Charnes, A., Cooper, W., Lewin, A. Y., & Seiford, L. M. (Eds.). 1994. Data Envelopment
Analysis: Theory, Methodology and Applications. Boston: Kluwer Academic Publishers.
Cheng, J. L. C. & Kesner, I. F. 1997. Organizational Slack and Response to

Environmental Shifts: The Impact of Resource Allocation Patterns. Journal of
Management, 23(1): 1-18.
Cherchye, L. 2001. Using data envelopment analysis to assess macroeconomic policy

performance. Applied Economics, 33: 407-416.
Chesbrough, H. W. 2000. Designing Corporate Ventures in the Shadow of Private

Venture Capital. California Management Review, 42(3): 31-49.
Chesbrough, H. W. & Socolof, S. J. 2000. Creating New Ventures from Bell Labs
Technologies. Research-Technology Management, 43(2): 13-17.

Review, 80(3): 4-11.

Coase, R. 1937. The Nature of the Firm. Economica N.S.(4): 386-405.
Coelli, T. 1996. A Guide to DEAP Version 2.1: A Data Envelopment Analysis

(Computer) Program.
Cohen, W., Levin, R. C., & Mowery, D. C. 1987. Firm Size and R&D Intensity: A Re-
examination. The Journal of Industrial Economics, XXXV(4): 543-565.

Cook, P. & Uchida, Y. 2002. Productivity growth in east Asia: a reappraisal. Applied
Economics, 34: 1195-1207.
Delmas, M. & Tokat, Y. 2005. Deregulation, Governance Structures, and Efficiency: The
U.S. Electric Utility Sector. Strategic Management Journal, 26(5): 441-460.
Dierickx, I. & Cool, K. 1989. Asset Stock Accumulation and Sustainability of

Competitive Advantage. Management Science, 35(12): 1504-1510.
55
Durand, R. & Vargas, V. 2003. Ownership, organization, and private firms' efficient use
of resources. Strategic Management Journal, 24(7): 667-675.



Policy, 34(5): 615-639.

Eisenhardt, K. M. 1985. Control: Organizational and Economic Approaches.

Management Science, 31(2): 134-149.
Eisenhardt, K. M. 1989. Agency Theory: An Assessment and Review. Academy of

Emrouznejad, A. 2003. An alternative DEA measure: a case of OECD countries. Applied

Economics Letters, 10(12): 779-782.

EVCJ. 2002. Corporate Venturing Falls Off. European Venture Capital Journal(October):
53.
Fare, R., Grosskopf, S., Norris, M., & Zhang, Z. 1994. Productivity Growth, Technical
Progress, and Efficiency Change in Industrialized Countries. American Economic
Review, 84: 66-83.
Fredriksen, O. & Klofsten, M. 2001. Venture Capitalists' Governance of Their Portfolio

Companies. Journal of Enterprising Culture, 9(2): 201-219.

Paper # 6725.
56
Quarter): 1-17.
Greene, W. H. 2003. Econometric Analysis: Prentice Hall.
Hambrick, D. C. & MacMillan, I. C. 1985. Efficiency of Product R&D in Business Units:

The Role of Strategic Context. Academy of Management Journal, 28(3): 527-547.
Hellmann, T. 1998. Some Thoughts on the Theory of Corporate Venture Investing,

Corporate Governance Today: 37-43: Columbia Law School.

Economics, 64(2): 285-314.
Jensen, M. & Meckling, W. 1976. Theory of the Firm: Managerial Behavior, Agency
Costs, and Ownership Structure. Journal of Financial Economics, 3: 305-360.
Keil, T. 2004. Building External Corporate Venturing Capability. Journal of Management

Studies, 41(5): 799-825.
Keil, T., Zahra, S. A., & Maula, M. V. J. 2004. Explorative and Exploitative Learning
from Corporate Venture Capital: A Model of Program Level Determinants. Academy of
Management Best Conference Paper: L1-L6.
Kim, J. W. & Lee, H. K. 2004. Embodied and disembodied international spillovers of

R&D in OECD manufacturing industries. Technovation, 24: 359-368.
Lawson, M. B. 2001. In Praise of Slack: Time is of the Essence. Academy of

Management Executive, 15(3): 125-135.
Leonard-Barton, D. 1995. Wellsprings of Knowledge: Building and Sustaining the

Sources of Innovation. Boston: Harvard Business School Press.
Lerner, J. 1994. The Syndication of Venture Capital Investments. Financial Management,

23(3): 16-27.
Lerner, J. 1995. Venture Capitalists and the Oversight of Private Firms. Journal of
Finance, 50(1): 301-318.
Levin, R. C., Klevorick, A. K., Nelson, R. R., & Winter, S. G. 1987. Appropriating the
Returns from Industrial R&D. Brookings Papers on Economic Activity: 783-820.
Maidique, M. A. & Hayes, R. A. 1984. The Art of High Technology Management. Sloan
Malmquist, S. 1953. Index numbers and indifference surfaces. Trabajos de Estatistica, 4:

209-242.
57
McConnell, C. R. & Brue, S. L. 2002. Macroeconomics: Principles, Problems, and

Policies (15 ed.): McGraw-Hill Companies.
Penrose, E. T. 1959. The Theory of the Growth of the Firm. New York: Wiley.

Red Herring. 2005. VC Fundraising Up 62%. U.S. Venture Firms Raised More in the
First Three Quarters of 2005 than All Last Year. Last accessed on December 23, 2005 at
http://redherring.com/Article.aspx?a=14016&hed=VC%20Fundraising%20Up%2062%.
Riyanto, Y. E. & Schwienbacher, A. 2005. The Strategic Use of Corporate Venture

Financing for Securing Demand. Working Paper.
Sah, R. K. & Stiglitz, J. E. 1987. The Invariance of Market Innovation to the Number of
Firms. RAND Journal of Economics, 18(1): 98-108.
Schumpeter, J. A. 1934. The Theory of Economic Development. Cambridge, MA:

Harvard University Press.
Brothers.
Shane, S. 2001. Technology Regimes and New Firm Formation. Management Science,
47(9): 1173-1190.
Singh, J. V. 1986. Performance, Slack, and Risk Taking in Organizational Decision

Making. Academy of Management Journal, 29(3): 562-585.
Thursby, J. G. & Thursby, M. C. 2002. Who Is Selling the Ivory Tower? Sources of
Growth in University Licensing. Management Science, 48(1): 90-104.
Tripsas, M. 1997. Surviving Radical Technological Change Through Dynamic

Capability: Evidence from the Typesetter Industry. Industrial and Corporate Change,
6(2): 341-377.
von Hippel, E. 1994. "Sticky Information" and the Locus of Problem Solving:
Implications for Innovation. Management Science, 40(4): 429-439.
Wiggins, V. 2001. Comparing xtgls with regress, cluster(). Stata FAQ resources, last
accessed on June 17, 2005 at http://www.stata.com/support/faqs/stat/xtgls_rob.html.
58
Williamson, O. 1985. The Economic Institutions of Capitalism. New York: Free Press.
Wooldridge, J. M. 1999. Introductory Econometrics: A Modern Approach. Cincinnati:

South-Western College Pub.
Yeoh, P.-L. & Roth, K. 1999. An Empirical Analysis of Sustained Advantage in the U.S.
Pharmaceutical Industry: Impact of Firm Resources and Capabilities. Strategic
Management Journal, 20(7): 637-653.
Zahra, S. A. 1996. Governance, Ownership, and Corporate Entrepreneurship: The

Moderating Impact of Industry Technological Opportunities. Academy of Management
Journal, 39(6): 1713-1735.
Zahra, S. A., Schulze, W., & Anokhin, S. 2005. Types of CVC Deals and Corporate
Innovation. Working Paper.
Zajac, E. J., Golden, B. R., & Shortell, S. M. 1991. New Organizational Forms for
Enhancing Innovation - the Case of Internal Corporate Joint Ventures. Management
Science, 37(2): 170-184.
59
Figure 1. Intertemporal Production Frontier in Computer and Electronic Product
CLRT -1998
Assets utilization per unit of sales
WAVX -1998
WAVX -1999
PANL -2000
PANL -2001 WAVX -2002
CLRT -2000
3OPTI-2000
3OPTI-2002
3CBEX-2003
3CBEX-2002
Labor utilization per unit of sales
60
61
62
Figure 2. Interaction of CVC, R&D and organizational slack
High Slack Environment. CVC Intensity as Moderator
Rates of Corporate Innovation 200.00
150.00
100.00
50.00
0.00
-1 SD Mean +1SD
R&D Intensity
Low CVC Intensity High CVC Intensity
Low Slack Environment. CVC Intensity as Moderator

Rates of Corporate Innovation
200.00
150.00
100.00
50.00
0.00
-1 SD Mean +1SD
R&D Intensity
Low CVC Intensity High CVC Intensity
63
Appendix 1
This paper employs a classical approach to defining the frontier (Caves et al.,
1982; Charnes, Cooper, Lewin, & Seiford, 1994; Cherchye, 2001; Coelli, 1996; Cook &
Uchida, 2002; Emrouznejad, 2003; Fare et al., 1994). In principle, basic DEA model
where there are u=1,…,U DMUs using n=1,…N inputs xu,t in stage s to produce yu,t in
period t=1,…,T may be expressed by a programming problem of the following form:
U U
D u′, t ( x u′, t , y u′, t ) −1 = Maxθ u′ , θ u′ y u′, t ≤ ∑ z u , t y u , t , ∑z
u ′, t
≤ xn , and z u , t ≥ 0 ,
u, t u, t
xn
u =1 u =1
where θ u′ is a measure of the distance of u ′ from the frontier. If 1/ θ u′ =1, then u ′ lies on
the frontier; otherwise, u ′ lies interior to the frontier, and 1/ θ u′ represents the fraction of
possible output produced by u ′ (for more details see Appendix A at (Thursby et al.,
2002)). Mathematically, Malmquist decomposition may be represented as follows :
m( x u′, t +1 , y u′, t +1 , x u′, t , y u′, t ) =
1
⎡⎛ D u′, t +1 ( x u′, t +1 , y u′, t +1 ) ⎞⎤ ⎡⎛ D u′, t ( x u′, t +1 , y u′, t +1 ) ⎞⎛ D u′, t ( x u′, t , y u′, t ) ⎞⎤ 2
⎢⎜⎜ u ′, t u ′, t u ′, t
⎟⎟⎥ x ⎢⎜⎜ u′, t +1 u′, t +1 u′, t +1 ⎟⎟⎜⎜ u′, t +1 u′, t u′, t ⎟⎟⎥ . Here the first
⎣ ⎝ D ( x , y ) ⎠⎦ ⎣⎝ D (x ,y ) ⎠⎝ D ( x , y ) ⎠⎦
bracketed term is essentially a ratio of the efficiency measure θ u′ in period t to θ u′ in
period t+1; it is the component of productivity change due to movement toward (away
from) frontiers in periods t and t+1, or the growth due to catching up (lagging). It tends to
capture the effects of capacity utilization, differences in the structure of economy and
diffusion of technology . The second bracketed term is the component of productivity
change due to frontier shift between t and t+1. It is this term that represents technical
change, or rates of corporate innovation. Generally, values of technical change exceeding
64
1 indicate that the technology employed by a company in a particular year is superior to
the technology used in the previous year.
65
Appendix 2
Table 3. Industry Classification
Industrial Subindustry (Malmquist) NAICS

group included
(4-digits
aggregation)
1 Utilities Utilities 2211, 2212
2 Manufacturing Food and beverage manufacturing 3121, 3114
Paper manufacturing, printing, and related 3221, 3222,
support activities 3231
Pharmaceutical and Medicine Manufacturing 3254
Chemical Product and Preparation 3241, 3252,
Manufacturing 3236, 3259,
3261
Machinery Manufacturing 3332, 3333
Computer and Electronic Product 3340, 3341,
Manufacturing 3342, 3343,
3344, 3345
Miscellaneous Manufacturing 3353, 3391
Motor Vehicle Manufacturing 3361
3 Trade Wholesale Trade 4200, 4227,
4242
Miscellaneous Store and Nonstore Retailers 4532, 4541
4 Cargo Scheduled Air Transportation 4811
Couriers 4921
5 Information Newspaper, Periodical, Book, and Directory 5111
Publishers
Software Publishers 5112
Radio and Television Broadcasting, Cable and 5133, 5151,
Other Subscription Programming 5152
Telecommunications 5171, 5172
Internet Service Providers, Web Search Portals 5181, 5191
and Other Information Services
6 Other Health Care 6214, 6221
Other industries All else
66
67
Essay II. Making Sense of Corporate Venture Capital II:
Technology Push vs. Market Pull
ABSTRACT
By making equity investments into new ventures incumbent corporations pursue
multiple strategic objectives including ‘window on technology’ and securing demand on
the incumbents’ own products and services that stimulates scale efficiency gains. To date,
however, evidence of established firms being successful in their corporate venture capital
(CVC) investments is scarce. Building on the resource-based view of the firm and the
conceptual framework developed by Chesbrough this study attempts to answer the long-
standing question that motivates much of the CVC literature: when does a corporate
parent experience growth in the rates of corporate innovation and increase its scale
efficiency as a result of its CVC investments?
The analysis demonstrates that technological and market fit between the
corporation and the new venture affect the likelihood of fulfilling incumbents’ strategic
objectives in different ways. Market development potential – the ability of the startup to
create new or extend existing markets for the incumbent – is the key dimension. When
market development potential (MDP) is present, rates of corporate innovation increase
and scale efficiency gains follow. When MDP is combined with technological fit,
innovativeness soars. Without MDP, however, both rates of corporate innovation and
scale efficiency decline regardless of presence/absence of technological fit. These results
hold after controlling for organizational, industry, and temporal effects.
68
The study clarifies and contextualizes important findings reported by the CVC
literature and sheds light on some issues left unanswered by the prior research.
69
Making Sense of Corporate Venture Capital II: Technology Push vs. Market Pull
“Companies can justify VC investments if they

add value for their shareholders in ways that
the shareholders cannot do themselves. But
although companies might argue that their
core businesses give them superior knowledge
of technologies and markets and thus
advantages over other investors … evidence of
this is scarce.”
Henry Chesbrough, ‘Making Sense of

Corporate Venture Capital’
1. Introduction
Goals pursued by incumbent corporations making equity investments into new
ventures have evolved from mainly financial in early seventies to mainly strategic in late
nineties (Gompers & Lerner, 1998). As the early 2000s have been characterized by a
significant downward trend and high volatility in corporate venture capital (CVC)
activity (EVCJ, 2002b), strategic component is likely to become further magnified: in
times of crises strategically and not financially oriented corporations remain in the CVC
business (Aernoudt & San Jose, 2003; EVCJ, 2002a). Recent study of the CVC market
demonstrated that at least in some aspects economic volatility had little or no effect on
the strategic choices CVC programs made (Ernst & Young, 2002). In fact, some authors
suggest that depression periods are precisely the right timing for corporate
entrepreneurship (Burgelman, 1983). This is echoed by Reaume (2003) who suggests that
importance of CVC programs for the corporate parents may actually increase with
industry volatility.
70
Of the strategic goals the dominant is sourcing of innovation, or ‘window on
technology’ (Ernst & Young, 2002). Whether a corporation experiences an increase in
innovativeness as a result of its equity investments depends on the industry characteristics
(e.g., intellectual property protection), organization’s properties (e.g., absorptive
capacity), CVC unit governance system (i.e., autonomy), and other factors (Birkinshaw &
Hill, 2003; Dushnitsky & Lenox, 2005b). How CVC interacts with internal research and
development, and how this interplay affects corporate innovation is a function of
organizational slack: when slack is abundant CVC and R&D reinforce each other; when
it is scarce the relation is opposite (Anokhin & Schulze, 2006a).
Another determinant of success of absorbing by the incumbent of the know-how
inherent in new ventures’ ideas is the type of CVC deals themselves. Chesbrough (2002),
for instance, develops a framework that suggests that the new ventures supported by
corporate venture capital differ with respect to similarity of their technological
capabilities to those of the corporate parent and in terms of their ability to increase the
sales and profits of the incumbent’s products and services. To date, however, the effects
of different types of CVC investments on corporate innovation remain undocumented;
empirical findings are few and far between. Although elements of Chesbrough’s
framework could be seen in much of current CVC research (Dushnitsky & Lenox,
forthcoming; Riyanto & Schwienbacher, 2005) whether mainstream strategy literature is
capable of predicting the relationship between different kinds of investments and the rates
of corporate innovation remains to be tested.
Corporate innovativeness notwithstanding, incumbent firms may have other
strategic goals on their agenda such as accessing new markets and otherwise stimulating
71
demand on own products/services (Dushnitsky & Lenox, 2005a; Ernst & Young, 2002;
Gompers et al., 1998). With few exceptions these goals of CVC investments remain in
research limbo; existing work is limited to economic modeling and provides no empirical
support of the arguments (Riyanto et al., 2005). Where should corporations invest if their
goal is achieving scale efficiency gains and not necessarily technical advancement? How
should such investments differ from those that aim at increasing corporate
innovativeness?
It appears that empirical CVC literature is plagued with the unstated assumption
that all investments committed by the corporation are homogenous: typically, CVC
disbursements over a specified period are combined, and entire CVC programs are
somehow dichotomized1 (e.g., “active/passive”, “financial/strategic”) (Chesbrough &
Tucci, 2004; Dushnitsky et al., forthcoming). While there is a lot to be learned from such
studies, inasmuch as nothing precludes corporations from making investments of
different types, such aggregation is hardly justifiable: different types of deals may have
different strategic implications, and corporations often do pursue multiple goals when
initiating their CVC programs (Ernst & Young, 2002). Naturally, some investments are
likely to bring about technical advancement and innovation while others are more likely
to stimulate demand for the corporation’s products and thus lead to scale efficiency gains
rather than affect corporate innovation per se (Riyanto et al., 2005).
We believe that by taking into account different types of CVC investments and
avoiding somewhat crude dichotomies with respect to CVC this study may provide
1
For a rare exception see Dushnitsky (2004) who attempts to label particular investments and not programs
as substitutive and complementary
72
empirical evidence – which to date is scarce, as the quote from Chesbrough’s work at the
beginning of this paper suggests – to the acclaimed ability of the corporations to add
value to the shareholders by means of CVC above and beyond mediocre financial returns
(Gompers, 2002) due to the incumbents’ superior knowledge of technologies and
markets. This study synthesizes the literature on the strategic component of CVC and
explores the effects of the composition of CVC investment portfolio on the rates of
corporate innovation and scale efficiency gains.
The study proceeds as follows. Section 2 overviews strategic objectives of
corporate venture capital and briefly summarizes the resource-based view of the firm
(RBV), which conceptually should be capable of explaining the outcomes of CVC deals
of different types. Section 3 introduces an adaptation of Chesbrough’s classification
scheme and applies RBV to this framework to formulate testable hypotheses with respect
to the effect of different types of CVC deals on corporate innovation and scale efficiency
gains. Section 4 explains methodological aspects of the study, which is followed by the
results (Section 5). Section 6 discusses the results and tries to explain the findings
inconsistent with the mainstream strategy expectations. The study concludes with the
implications for scholars and policy makers, and suggests limitations and directions for
future work.
2. Literature Review
2.1. Strategic Objectives of CVC Investments
Corporate venture capital literature reports a plethora of strategic objectives
pursued by the corporations through CVC investments including ‘window on
73
technology’, leveraging internal technological developments, importing/enhancing
innovation with existing business units, corporate diversification, securing demand on
own products, searching acquisition targets, tapping into foreign markets, etc.
(Dushnitsky et al., forthcoming; Ernst & Young, 2002; Gompers, 2002; Riyanto et al.,
2005). There are several approaches to CVC programs classification. Aernoudt et al.
(2003), for instance, suggested classifying them along internal/external and direct/indirect
dimensions. Chesbrough (2002) offered strategic/financial orientation and tight/loose
technological link dimensions along which to group CVC investments. The programs
could also be grouped by their goals: some aim at innovation, or technical advancement
while others target corporate sales and as such are more likely to bring about scale
efficiency gains. It appears, though, that unambiguous classification of corporate VC
programs may be somewhat problematic since “espoused” goals of (corporate) venture
capitalists may differ from those “in use” (Aernoudt et al., 2003; Christopher, 2000;
Hardymon, DeNino, & Salter, 1983; Shepherd, 1999). This is true even for the most
intuitively appealing strategic/financial classification: while some studies report that
majority of the corporations put less emphasis on investing money chiefly to get a return
(EVCJ, 2002a), and 93% of CVCs list strategic objectives as one of their main objectives
(Hellmann, 2002), other studies report precisely the opposite – namely, growing focus on
financial returns and not a strategic fit with the parent company’s business (Christopher,
2000). Besides, as said above, nothing precludes corporations from making multiple
investments of different types. Thus, to study strategic outcomes of CVC investments it is
desirable to concentrate on the characteristics of particular deals and not aggregate CVC
programs.
74
Generally, for strategic benefits to realize there has to be fit of some sort between
the investees’ and the corporate parent’s products, markets and/or knowledge base.
Accordingly, the concept of fit which plays a central role in strategic management
literature (Venkatraman, 1989; Venkatraman & Camillus, 1984) seems to be equally
important in the field of corporate venture capital. “Fit” is often used as a synonym to
words like “match”, “coalignment”, “congruence”, and “complementarity” among others
(Schoonhoven, 1981). Venkatraman et al. (1984) classified approaches to conceptualizing
fit into “content of fit” and “pattern of interaction” schools of thought. Content-based
school focuses on strategic actions to be taken to match different environmental
conditions. Pattern-based conceptualization of fit focuses on the process of strategy
making; strategy itself is an ongoing pattern of matching the different elements – some
within the organizational boundaries (competences and resources) and others dealing
with the environment (opportunities and threats) (Venkatraman et al., 1984: p. 514).
Unlike content school which is mostly devoted to organization-environment type of
theorizing, pattern school implies interorganizational relationships.
Chesbrough’s view of ‘fit’ and ‘strategy’ in CVC context that we build on in this
paper is somewhat different and circumscribed rather narrowly. First dimension of his
CVC classification framework is strategic/financial. ‘Strategic’ in Chesbrough’s work
refers to investments made primarily to increase the sales and profits of the corporation’s
own business. Thus, it appeals to one particular strategic goal described by the prior CVC
literature: stimulating demand on the corporation’s products (and resulting scale
efficiency gains). The other main strategic objective of CVC investments – window on
technology – is covered by the second dimension of Chesbrough’s framework – the
75
degree to which companies in the investment portfolio are linked to the investing
company’s current operational capabilities. Included in the Chesbrough’s notion of
operational capabilities are the incumbent’s resources and competencies (processes).
Overall, the broad Venkatraman’s conceptualization of fit and the narrower
Chesbrough’s approach clearly identify several important elements defining the corporate
parent-new venture dyad: matching competencies and resources, dealing with
opportunities and threats, and developing interorganizational relationships. Through CVC
investments corporations identify and establish control over promising technologies, or
exploit opportunities. Corporate venture capital helps the corporation enter alternative
markets thus diminishing its dependency on a narrowly defined competitive environment.
In other words, corporate venturing may be instrumental in neutralizing environmental
threats. CVC may also lead to a better usage of assets already controlled by the
corporation. Taken together, that qualifies corporate venture capital as a resource
explains Barney (1991). The resource-based view of the firm, accordingly, could be
expected to provide valuable insights into how competencies and resources of the
corporate parent and the new venture have to be matched to add value – in terms of
innovation or scale efficiency gains – to the incumbent firm.
2.2. The Resource-based View of the Firm
Under RBV, the firm’s ability to make better use of its resources (Penrose, 1959)
or to achieve more with a comparable resource bundle leads to superior performance
typically manifested in the above normal returns (Barney, 1991; Conner, 1991; Dierickx
& Cool, 1989; Wernerfelt, 1984). The resource-based perspective attributes sustainable
76
competitive advantage to superiority of resources under the firm’s control. Resources
include all assets, capabilities, organizational processes, firm attributes, information,
knowledge, etc. controlled by a firm that enable the firm to conceive of and implement
strategies that improve its efficiency and effectiveness (Barney, 1991, p. 101, emphasis
added). This definition is important: a resource does not have to be owned by the
corporation to enable superior performance; it has to somehow be controlled. Corporate
venture capital provides a mechanism to control a resource through equity investments:
over 85% of the corporations that have policies on taking seats on the boards of the new
ventures they support do take advantage of this opportunity (Anokhin & Schulze, 2006b).
As such, they are actively involved in strategic steering of the new venture by supervising
the startups’ management. They also provide technical assistance to the new ventures and
perform the due diligence of their business plans. Accordingly, corporate parents get an
accurate idea of the startups’ technological know-how, and may in some ways affect the
new ventures’ development trajectories.
To ensure superior outcomes, the resources need to be valuable, rare, imperfectly
imitable, and non-substitutable (Barney, 1991). Firm attributes are valuable resources
when they exploit opportunities or neutralize threats in a firm’s environment. They are
rare when a strategy based on their use cannot be simultaneously implemented by a large
number of other firms. This is true for the bundles of resources as well. To have a
potential of generating a competitive advantage, the number of firms that possess a
resource or a bundle of resources has to be less than the number of firms needed to
generate perfect competition dynamics in an industry (Barney, 1991). Firm attributes are
imperfectly imitable when there are unique historical conditions, causally ambiguous
77
processes and social complexities involved in their acquisition or development (Barney,
1991; Dierickx et al., 1989). They are non-substitutable when there are no strategically
equivalent resources (or bundles of the resources) that are themselves either not rare or
imitable (Barney, 1991). Overall, resources needed to establish a sustainable competitive
advantage cannot simply be bought; Dierickx et al. (1989) call this nontradeability. To
obtain such resources the firm needs to possess superior information, be lucky, or both
(Dierickx et al., 1989). Leaving luck aside, the implication is that corporate venture
capital investments have higher success potential when done in the fields where the
corporation is likely to possess superior knowledge. This is not to say that such
investments have to be committed to the exactly same niche that the corporation is in;
what matters is the complementarity and/or transferability of the competencies between
the corporation and the investee (in both directions).
A firm’s capacity to deploy (a bundle of) resources using organizational processes
is known as capabilities (Amit & Schoemaker, 1993; Grant, 1991) which in essence is a
second-order resource. Capabilities could generally be defined as a high-level routine (or
collection of routines) that, together with its implementing input flows, confers upon an
organization’s management a set of decision options for producing significant outputs of
a particular type (Winter, 2003: p. 991). Along with assets, organizational processes, firm
attributes, information, knowledge, etc., capabilities are a part of the firm’s resource
bundle (Barney, 1991). Capabilities are always objective-specific; there are no “general
purpose” capabilities (Winter, 2003). Depending on the objective, one may differentiate
technological, marketing (Verona, 1999), organizational (Collinson, 2001), integrative
(Henderson, 1994; Lawrence & Lorsch, 1967; Yeoh & Roth, 1999), combinative (Kogut
78
& Zander, 1992), dynamic capabilities (Teece, Pisano, & Shuen, 1997), dynamic
technical capabilities (Tripsas, 1997), etc. The latter (integrative, combinative, dynamic)
govern the rate of change of the former (which are referred to as ordinary or first-order)
capabilities and thus could be considered second-order capabilities (Collis, 1994; Winter,
2003).
Of the second-order ones, integrative capability – or the ability of the firm to
integrate knowledge across both firm and disciplinary boundaries (Henderson, 1994) – is
particularly important in the CVC context. Corporate venturing requires companies to
develop and manage specialist knowledge and expertise which is distributed across
different divisions internally and among their outside partners including suppliers,
customers, specialist R&D organizations, etc. (Collinson, 2001). Different aspects of the
fundamental idea behind the notion of integrative capabilities have also been captured by
the concepts of architectural competence (Henderson & Cockburn, 1994), combinative
capabilities (Kogut et al., 1992), dynamic capabilities (Teece et al., 1997) and related
categories. Another related category is absorptive capacity, which has been shown to
positively affect the ability of the corporate parent to learn from its CVC investments
(Dushnitsky et al., 2005a). Research performance is positively associated with the ability
to span the boundaries of the firm; unique abilities to redeploy existing knowledge may
be fundamental to long-term strategic advantage (Henderson et al., 1994; Kogut et al.,
1992). Integrative capabilities allow both effectively capturing the externally developed
ideas and combining them with existing knowledge in a productive and efficient manner.
As Henderson et al. (1994) theorize, other things being equal, firms with the ability to
encourage and maintain an extensive flow of information across the boundaries of the
79
firm and across the boundaries between (scientific) disciplines within the firm will be
more successful in terms of realized innovation.
3. Hypotheses Development
As noted above, to better understand corporate venture capital, Chesbrough
(2002) offered two dimensions on which CVC investments differ: operational link, or the
degree to which companies in the investment portfolio are linked to the investing
company’s current operational capabilities (resources and processes), and strategic fit, or
the degree to which investments of the corporate parent are likely to increase sales and
profits of its own business. Since this view of strategy diverges from the rest of the CVC
literature which is much broader in terms of what constitutes strategic benefits for the
corporate parent, we re-label it as market development potential to remain consistent with
prior CVC research. Combined, these two dimensions allow differentiating between four
different types of CVC investments: driving, enabling, emerging, and passive. Driving
are CVC investments characterized by a tight link of the new venture to the operational
capability of the corporation, and high market development potential. Enabling have high
market development potential but are only loosely linked to the corporation’s operational
capability. Emerging do not provide a market development potential although have tight
links to the corporate parent’s operational capabilities. Finally, passive investments
neither match operational capabilities of the corporation nor are capable of extending its
market presence. They are believed to be committed primarily for financial reasons and
are most similar to the investments made by independent venture capital firms
(Chesbrough, 2000: p. 6-10). This classification framework may be applied to
80
circumscribe multiple objectives pursued by the corporations through their CVC
programs. For instance, investments that aim at ‘window on technology’ could be defined
as driving, securing demand on own products – as enabling, leveraging internal
technological developments – as emerging, and so on.
As RBV informs us, resources required to establish a sustainable competitive
advantage cannot simply be bought. The potential value of the resource should be non-
obvious to market participants; otherwise, it would have been priced at the level where
above-normal returns were appropriated by the resource-owner, not the resource
purchaser. This rarely happens due to bounded rationality of the resource owners
(Utterback, 1974). CVC investments may generate competitive advantage when they
complement other resources controlled by the corporation in a synergy-generating, non-
obvious way. Same processes that make firm attributes imperfectly imitable (unique
historical conditions, causally ambiguous processes and social complexities involved in
their acquisition or development) make the fit and the resulting resource bundle value
generation capacity non-obvious to the market (Barney, 1991; Dierickx et al., 1989) and
thus allow benefits appropriation by the corporate parent. This may happen when the
corporation has superior knowledge about new venture’s potential and is more likely to
take place in case of strong technological or market overlaps between the corporate
parent and the startup.
Driving investments have a strong link to the corporate parent’s operational
capabilities. For instance, by investing into computer software and services startups
Adobe is likely to build relationships with companies that not only may become its
customers but may also possess know-how that is of strategic interest to Adobe itself. By
81
monitoring its investee’s activities, Adobe is likely to internalize and absorb some of the
new venture’s know-how and advance its own technologies. Emerging investments,
while not necessarily opening new or extending existing markets for the incumbent
corporation, still possess resources and competencies similar to those of the corporate
parent, and are thus likely to result in synergies and technological advancement. Thus,
investments by DaimlerChrysler into computer hardware manufacturing are unlikely to
directly affect demand on the parent company’s main product (automobiles) but perhaps
automobile’s certain systems – such as electronic control unit – may benefit from these
investments.
When operational capabilities match, technical advancement may be expected to
emerge as a result of the investment. As the innovation literature informs us, even when
appropriability regimes are tight knowledge transfer and spillover occur (Shane, 2001).
When it is weak, absorption by the corporation of the innovation inherent in new
venture’s ideas is especially effective (Dushnitsky et al., 2005a). In the case of CVC
investments such transfer is actively encouraged by the corporation (Keil, Zahra, &
Maula, 2004). Not only do the corporations provide technical assistance to new ventures
they support and delegate their representatives to the new ventures’ boards (Anokhin et
al., 2006b), but they also perform the due diligence of the new ventures’ business plans
which invokes disclosure of valuable information. Dushnitsky (2004) in his study of the
likelihood of formation of the ‘corporation-new venture’ pairs explains how such
disclosure may expose the proprietary technologies of the new venture and lead to the
know-how leak to the incumbent. Accordingly, driving and emerging investments that
82
have tight link to the operational capabilities of the corporation are likely to result in
increase of the corporation’s innovativeness (technical advancement). Therefore:
Hypothesis 1. Driving CVC investments are positively associated with the rates of
corporate innovation.
Hypothesis 2. Emerging CVC investments are positively associated with the rates of
corporate innovation.
Apart from providing tight operational capabilities link, driving investments
imply that a new venture is likely to open new or extend existing markets for the
incumbent corporation. Such is the nature of eBay’s investments into communications
and media. Corporate parent’s market presence may also be extended through
investments into new ventures that are likely to use the corporation’s products in their
day-to-day business even if they do little to enhance incumbent’s operational capabilities.
Chesbrough labeled them enabling investments. Glaxosmithkline’s support of healthcare-
related organizations may provide and example: although resources and competencies
here are drastically different, healthcare providers are the natural customers for the
Glaxosmithkline’s products. When market development potential of the CVC investment
is high, one could expect such investment to stimulate demand for the corporation’s own
products and services. Accordingly, the following should be true for both driving and
enabling investments:
Hypothesis 3. Driving CVC investments are positively associated with the rates of scale
efficiency gains.
83
Hypothesis 4. Enabling CVC investments are positively associated with the rates of scale
efficiency gains.
Finally, investments in startups that have no link to the operational capability of
the corporate parent and could not extend its market presence are believed to be done for
financial and not strategic reasons and are thus a part of Chesbrough’s passive category.
While they are expected to affect the corporation’s bottom line, existing literature is
inconclusive as to how they should influence corporate innovation or scale efficiency
gains: although in some cases “learning by doing something else” could happen
(Schilling, Vidal, Ployhart, & Marangoni, 2003) due to lack of fit such investments are
unlikely to lead to technical advancement or efficiency improvement.
4. Data and Method
4.1. Database Construction
To test proposed relationships we construct a unique dataset by matching multiple
secondary data sources2. VentureXpert by Venture Economics and Corporate Venturing
Directory & Yearbook (hereafter – the Yearbook) by AssetAlternatives are utilized to
reconstruct the pattern of CVC investments by incumbent corporations. While both data
sources have been used in CVC research extensively (Birkinshaw et al., 2003;
Chesbrough et al., 2004; Dushnitsky, 2002, 2004; Keil et al., 2004; Maula & Murray,
2000) they are known to have certain deficiencies: VentureXpert may inflate the number
of investment rounds (Lerner, 1994, 1995) and the Yearbook may double count particular
deals (Anokhin et al., 2006a). Besides, each data source has information on some deals
that the other database does not provide. Thus, by carefully matching the data we are able
2
84
to obtain the most accurate information on the CVC disbursements of the corporations.
We only consider investments committed during 1998-2001 as this period is best covered
by both databases. U.S. Census Bureau’s NAICS and Bureau of Economic Analysis’
1998-2004 Annual Input-Output tables are used to classify CVC investments into driving,
enabling, emerging, and passive.
After matching the data on CVC deals reported by VentureXpert and the
Yearbook, we merge our database with the annual firm-level accounting and financial
data from Standard & Poor’s Compustat. Since the data reported in the Compustat relate
to a financial and not a calendar year of the corporation, we do not use annual aggregates
reported by VentureXpert directly but rather look at the exact dates of particular deals to
match them to appropriate financial years. Thus, for a corporation with financial year
starting in March and ending in February we would consider a CVC investment made in
January of 2000 as a part of year 1999. To test different time lags, we look at financial
and accounting data over the six-year period of 1998-2003. The merger of VentureXpert,
the Yearbook, and Compustat yields a sample of 163 corporations that engaged in
corporate venture capital investments during years 1998 to 20013.
4.2. Dependent variables: inferring the rates of realized corporate innovation
and scale efficiency gains
Under the resource-based view of the firm, superior performance can be attributed
to the improved use of the firm’s resources when more is achieved with a comparable
3
We deliberately excluded certain industries such as financial services, real estate, hotels, etc., and
companies for which it is not possible to designate a primary industry affiliation (e.g., General Electric)
85
resource bundle.4 Improvement in the deployment of the resources may be achieved in
two ways: by creating a new way to combine resources (that is, advancing technology, or
innovating) and by increasing efficiency of resource usage under already existing
technologies without innovation per se (through, for instance, scale efficiency gains)
(Fare, Grosskopf, Norris, & Zhang, 1994; Penrose, 1959; Thursby & Thursby, 2002). The
former corresponds to the proclaimed CVC strategic goal of obtaining a ‘window on
technology’ while the latter reflects attempts to secure demand on the corporation’s own
products, tap into foreign markets, etc. with the help of the CVC.
There exists a technique that allows us to use secondary data from Compustat to
compute total factor productivity change and decompose it into technical change
(evidence of realized corporate innovation) and efficiency change, and distill scale
efficiency change from the latter. The technique has been known in operations research
for over 50 years (Malmquist, 1953) and has recently been adopted by economics (Fare et
al., 1994), innovation (Thursby et al., 2002) and strategy literatures (Durand & Vargas,
2003). The idea behind the technique is as follows. Corporations that belong to the same
industry combine homogenous inputs (such as labor and capital) in different ways to
produce comparable outputs (represented by sales). Some corporations use less labor than
others, some use less capital, while yet others use relatively large quantities of labor and
capital to produce the same amount of sales. Companies with best combinations of inputs
and outputs define the best available technology at the time; they are said to determine
the production frontier. Over time, as technology advances, frontier shifts.

4
The notion of comparability does not imply that resources are truly homogenous. In fact, there always
exists some heterogeneity, and “comparable” here refers to markets’ homogenous pricing of heterogeneous
inputs and outputs.
86
The technique known as Malmquist productivity index decomposition tracks
relative positions of different companies from year to year in terms of sets of their inputs
and outputs. It monitors the movement of less effective companies toward the frontier
(which does not require innovation and suggests that firms are simply becoming more
efficient at something already known) and shifts of the frontier itself (caused by technical
advancement, or introduction in year t by some companies of technologies superior to
those employed in year (t-1)). The former component – efficiency change – is
decomposable into scale efficiency change and pure efficiency change. The latter one –
technical advancement – is accepted in the literature as evidence of realized innovation
(Fare et al., 1994).
Essentially, the technique makes a comparison of sets of inputs and outputs for
different corporations within their subindustries in consecutive time periods, infers from
this total factor productivity change and by decomposing it allows to estimate the rates of
corporate innovation and scale efficiency gains. Mathematically the technique is
explained in Thursby et al. (2002) and Fare et al. (1994). The programming algorithm is
realized in DEAP software (Coelli, 1996; Hollingsworth, 2004). Overall, we analyze over
12,000 firm-year observations to arrive at estimates of technical advancement and scale
efficiency change5. We run the analyses separately for 20 subindustries to ensure
comparability of the firms and meaningfulness of the frontier in each run. Figure 3
illustrates the dynamics of the frontier based on the data for the computer and electronic
product manufacturing industry for years 1998-2003. We use the estimate of technical
5
For each run, we include all active firms with no missing values reported by Compustat operating in a
given subindustry to obtain the most accurate estimate of the technology prevalent in that industry in each
year.
87
advancement as our measure of the rates of corporate innovation (RCI). Values of RCI
exceeding 100 suggest that the technology employed by the corporation is superior to the
one used in the previous period. We believe that RCI is superior to other measures of
innovation such as patents in that it reflects actual and not potential innovation, and is
more general than patents which could only be assigned for a narrowly defined set of
ideas and are not always acted upon. The estimate of scale efficiency change is our
measure of scale efficiency gains (SEG). Again, values of SEG exceeding 100 indicate
that the corporation has achieved scale efficiency gains as compared to the previous
period. Compustat is employed to collect information about labor (employee count),
corporate assets, and corporate sales.
4.2. Independent variables: classifying CVC investments into driving,
enabling, emerging, and passive
To classify all CVC deals into four groups we employ a multistep algorithm.
First, we combine information contained in both VentureXpert and the Yearbook about
CVC deals committed during the specified period. Of particular interest to us is the
startup’s industry. Based on VentureXpert’s classification and the Yearbook’s description
of the new venture and its industry, we assign NAICS codes to the new ventures. We then
compare the corporate parent’s NAICS code to that of the new venture, and in case of a
match (at two to four digits aggregation) assign the deal a value of 1 on the operational
capabilities link (OCL) dimension and 0 otherwise. Similar approach has been used by
88
Dushnitsky (2004) to determine whether or not a CVC investment is likely to be a
substitute to the corporate parent’s line of products.
Next, we explore the Bureau of Economic Analysis’ Input-Output tables. We
record how much of the corporate parent industry’s output is consumed by CVC
investees’ industries. We then classify the new ventures as having high market
development potential for the corporate parent if their industries account for a certain
portion of this consumption. Two cut-off values are used: 10% and 5% of the average
consumption. Results reported in the next section of the paper utilize the first threshold.
Then, we assign a CVC investment a value of 1 on market development potential (MDP)
dimension if its industry consumes at or above the cut-off value and 0 otherwise. It
indicates whether or not the corporation’s investment in the new venture is likely to
generate noticeable increase of demand on the corporate’s offerings. Dushnitsky (2004)
employed similar approach to determine whether or not a particular investment is likely
to be complementary to the parent corporation’s line of business.
Our final step – classification of CVC investments into four groups – departs from
Dushnitsky’s approach. Unlike Dushnitsky and similarly to Chesbrough we believe that a
CVC investment may combine the two properties – operational capabilities link and
market development potential – in a meaningful way. If the deal stands high on both OCL
and MDP dimensions, it is classified as driving. If the value on the OCL scale is 1 and on
the MDP scale is 0, it is classified as emerging. Investments with 0 on the OCL scale and
1 on the MDP scale are considered enabling. Finally, investments with 0 on both
dimensions are considered passive. There remains a portion of investments that cannot be
classified since their industry affiliation is unclear. We run our analysis with and without
89
this group of deals, which does not affect our substantive results. Both sets of results are
reported.
We only look at the number of new ventures supported by the incumbent
corporation in a given year in each of the categories. There are three main reasons for
choosing this measure over a more intuitive dollar amount one. First, corporations tap
into the knowledge developed by the new venture regardless of the size of their
investment (Reaume, 2003). Second, the amount invested is often simply a function of
the investment round and is not indicative of the investment’s importance or relevance:
later rounds typically require more significant investments (Gompers et al., 1998). Third,
the Yearbook does not report amounts invested by a particular corporation in individual
deals and instead provides the overall figure. Although our data allow us to estimate the
dollar amount of those investments such estimation may introduce unnecessary noise. In
any case, the number of CVC investments is known to be highly correlated with the
dollar amount of CVC investments (Anokhin et al. (2006b) report a correlation
coefficient of 0.85).
4.3. Control variables
Integrative capability, or the ability of the corporation to integrate knowledge
across both firm and disciplinary boundaries (Henderson, 1994) may affect the rates with
which the corporation integrates external knowledge and combines it with existing
knowledge in a synergy-generating way. Thus, we control for the integrative capability of
the corporation. We operationalize it in a number of ways. First, we create a composite
score that takes into account the external component of integrative capability (the number
90
of distinct countries where the corporation invests its CVC money) and the internal
component (whether or not the CVC team has individuals with different functional
backgrounds, and whether or not it reports to one of the top managers of the corporation)
(Cockburn, Henderson, & Stern, 2000; Collinson, 2001; Henderson et al., 1994; Yeoh et
al., 1999). Information required to create this variable is derived from the Yearbook.
Second, we approximate integrative capability with the measure of corporate venturing
expertise (CVE) which we measure as the number of years since the corporation was first
mentioned in VentureXpert as a CVC investor or co-investor. Corporations with a long
history of participating in CVC activities are more likely to develop unique approaches to
identifying, supporting and assisting new ventures as well as appropriating the fair share
of the benefits based on the complementarities of the corporation-startup competencies.
Third, we use a binary variable seats, which takes on a value of 1 if the corporation takes
seats on the boards of the new ventures it supports with CVC and 0 otherwise. The idea is
that if the corporate parent assigns a representative to closely monitor the activities of the
new venture, it is more likely to recognize and internalize the innovative potential of the
new venture’s technology.
We account for the internal R&D intensity. Following Dushnitsky et al.
(forthcoming) we calculate it as the ratio of R&D outlays to the dollar amount of the
corporation’s assets rather than sales. Adjustment is made for the industry average as is
commonplace in the strategic management literature (Schilling, 2002). The data is
directly available from Compustat. We also control for the organizational slack generally
defined as a cushion of excess resources that can be used in a discretionary manner
(Bourgeois, 1981). There are several important reasons to control for slack. First,
91
potentially slack may determine both CVC and R&D spending: when firms have extra
resources, they might spend more on R&D and other things, such as CVC. When firms
have no extra resources, they may cut back on both (Chesbrough et al., 2004). In other
words, slack may affect the levels of both CVC and R&D. Second, when the firm is “lean
and mean”, competition for the resources is likely to become a factor in determining the
priorities of the corporation’s business units (Cheng & Kesner, 1997; Lawson, 2001) and
thus could affect the relationships between CVC and R&D. Prior research confirms that
availability of organizational slack affects the interplay of CVC and R&D (Anokhin et
al., 2006a). We operationalize organizational slack as the current ratio or the ability of the
firm to meet current obligations (current assets divided by current liabilities) (Singh,
1986). The data is directly available from Compustat.
Investment Risk is an important determinant of venture capital activity outcomes.
Corporate venture capital tends to invest in more established, less risky companies
(Anokhin et al., 2006a; Ernst & Young, 2002; Riyanto et al., 2005). Conventional R&D-
based measures of risk-taking may not be appropriate as high R&D spending may simply
reflect internal inefficiencies and high agency costs rather than aggressive risk-taking
attitude (Zahra, 1996). In line with the venture capital literature, we use the corporation’s
preferred investment round as a measure of risk (Ernst & Young, 2002; Fredriksen &
Klofsten, 2001). We dichotomize investments into relatively risky (seed and early stages)
and relatively non-risky (extension, later, and balanced stages). To create the variable we
use the data from both the Yearbook and VentureXpert. Since level of CVC activity in
the industry creates pressure to innovate (Zajac, Golden, & Shortell, 1991) we control for
industry membership. We also control for temporal effects (include dummy variable for
92
the year 2000 known for the crash of the dot.com market which constituted a significant
portion of the CVC investment portfolio) and firm size, operationalized as the log of
sales.
5. Results
Descriptive statistics can be found in Table 4. Of those investments that can be
classified unambiguously, on average corporations invest most actively into passive
deals: over 50% of CVC money goes into new ventures that provide neither operational
capabilities link nor market development potential to the incumbent. This is in line with
estimates provided in Chesbrough (2002). Over 16% of CVC funds are assigned to
driving investments, about one-fifth targets emerging deals, and less than 5% is invested
into ventures that have strong market development potential without providing
operational capabilities link. Corporations with active corporate venture capital programs
on average invest in 5.45 startups annually while the most active companies may support
over 250 new ventures per year (e.g., Intel in year 2000). On average, corporations had
4.41 years of CVC experience during the specified period. Only 15% of them had
strategies that could be defined as risky; the majority preferred to invest in later-staged
new ventures that already had a product.
The correlation coefficients are all within the tolerance limits with exception of
passive and driving investments which share about 64% of variance. Multicollinearity
testing, however, did not reveal problems, and all VIFs were under 2; accordingly,
passive investments were retained as a predictor variable for the analyses.
93
The results reported below were produced by feasible generalized least squares
estimator corrected for autocorrelation. FGLS is required to address possible violation of
the independence assumption by the RCI and SEG variables (Thursby et al., 2002).
Results of hypotheses testing are presented in Table 5. Models 1 through 3 have rates of
corporate innovation as a dependent variable (taken with a two-year forward temporal lag
to account for time necessary for the effect of the investments to be realized by the
corporation). They differ by the operationalization of the integrative capabilities: Model 1
uses composite score that takes into account external and internal components of
integrative capability; Model 2 utilizes CV expertise; and Model 3 employs the binary
variable ‘seats’. Models 4-6 are similar to Models 1-3 but exclude unclassified
investments to check the robustness of the estimates. Model 7 utilizes scale efficiency
gains as a dependent variable; Model 8 is a replica of Model 7 without unclassified
investments in the predictors set. All models demonstrate statistical significance at
p<0.001 level.
Results lend support to Hypothesis 1. Driving investments are positively
statistically significantly related to the rates of corporate innovation. Hypothesis 2 is
rejected. In fact, the relationship between emerging investments and rates of corporate
innovation is negative. Support by the incumbents of new ventures with related
technological capabilities but lack of market development potential is detrimental for
corporate innovation. This contradicts the predictions based on the resource-based view
of the firm. We discuss this intriguing finding in the following section of the paper.
94
Hypothesis 3 is supported. Driving investments are positively associated with scale
efficiency gains. Enabling investments also demonstrate positive relationship to scale
efficiency gains thus lending support to Hypothesis 4.
6. Discussion
First, our findings confirm that to understand implications of CVC for the
corporation it is desirable to study particular CVC deals, and not entire CVC programs.
Indeed, as Chesbrough suggested, different types of deals affect the corporate parent in
different ways. Driving investments – i.e., support by the incumbent of new ventures with
similar technologies and high market development potential – are most beneficial for the
corporation. They positively affect both the rates of corporate innovation and scale
efficiency gains. The fact that only one out of six CVC investments targets such ventures
is inconsistent with the espoused ‘strategic’ goals of the majority of CVC programs. This
finding is in line with research on independent venture capitalists investment practices
that demonstrates that espoused criteria are often different from those in use.
Enabling investments are beneficial for the corporation. Both corporate
innovation and scale efficiency stand to gain from investments in companies that provide
strong market development potential for the corporate parent even though they do not
match operational capabilities of the incumbent. Based on the magnitude of the
regression coefficients, positive effect of enabling investments on scale efficiency gains
is more than three times higher than the effect of driving investments. It is thus surprising
95
that enabling investments constitute less than 5% of an average CVC portfolio. While
such positive effect is expected in case of scale efficiency gains, it is somewhat surprising
for the rates of corporate innovation. It appears that pull from the market is at least as
important to achieve technical advancement as is push from the technological side.
Interestingly, driving investments which are positively related to corporate innovation are
characterized by strong market development potential.
Contrary to our expectations, emerging investments turn out to be detrimental for
the corporation in terms of both innovation and scale efficiency gains. Again, results are
not surprising in case of scale efficiency gains: after all, emerging investments have no
market development potential. What is surprising, though, is the negative effect of
emerging investments on the rates of corporate innovation. This seems to contradict our
predictions based on the resource-based view of the firm. Apparently, technology push –
i.e., surge of related technologies in the environment that the corporate parent controls –
by itself does not stimulate incumbent’s innovativeness. In this light, lack of significance
on all three alternative measures of integrative capabilities makes sense as well:
innovation is not in fact sourced and internalized from elsewhere, it is internally begotten,
and the rates of innovation production are particularly high when there is a strong market
call. CVC investments are especially strong predictors of the technical change when
technology push meets market pull: magnitudes of the regression coefficients suggest that
the effect of driving investments which combine both technology and market fit on the
rates of corporate innovation is more than three times as strong as the effect of enabling
investments that provide market fit alone.
96
Over half of the investments could be classified as passive: most of the new
ventures supported by CVC fit neither corporate parent’s technologies nor its markets.
Hence, Chesbrough’s comment on the lack of evidence for the ability of the corporation
to add value for their shareholders in ways that the shareholders cannot do themselves
quoted in the beginning of this paper should come at no surprise: corporations have no
superior information about the potential value of most new ventures they support. These
investments are negatively related to both corporate innovation and scale efficiency
gains. The finding is consistent throughout all eight models.
Inclusion/exclusion of unclassified investments does not change our substantive
results. In fact, coefficients demonstrate high robustness. By themselves unclassified
investments seem to be positively related to the rates of corporate innovation, although
statistical significance of this is marginal. Thus, we could expect that on average
unclassified investments provide some sort of fit with the corporate parent’s markets. At
the same time, their effect on scale efficiency gains is negative but lacks statistical
significance. We are uncertain how to interpret these findings, and leave it to future
research.
Investment risk does not affect scale efficiency gains but hampers the rates of
corporate innovation. Lack of significance for the relationship between R&D intensity
and innovation can be attributed to the effects of temporal lags: it has been shown that
unlike CVC effects that become visible with a two-year temporal lag, R&D effects are
realized quickly but are soon exhausted (Anokhin et al., 2006a). Consistent with prior
research, we see that organizational slack is important for corporate innovation. At the
same time, it is negatively related to scale efficiency gains. One of the explanations to
97
this fact may be that organizations without available underutilized capacity may
purposefully target scale efficiency and not technical advancement which requires
considerable resources as a means of improving their factor productivity. It is also of
interest that innovation is associated with larger firms while scale efficiency gains are
more of a smaller companies’ domain. This seems to be consistent with later
Schumpeterian views of innovation being introduced by large established companies;
smaller firms are better in adopting technologies developed by the industry leaders.
7. Conclusion
In this study we demonstrate that different types of CVC deals have different
implications for the corporate parent’s business: some investments are more likely than
others to increase the rates of corporate innovation and scale efficiency gains. To the best
of our knowledge, this is the first empirical attempt to apply conceptual framework
developed by Chesbrough to CVC investments committed by incumbent corporations.
We demonstrate that of the two dimensions on which new ventures may fit the corporate
parent market development potential is far more important than technological capabilities
link. Investments in startups that can create new or extend existing markets for the
corporation are likely to both result in corporate innovation and produce scale efficiency
gains.
While Dushnitsky and his colleagues’ work seems to indicate that corporate
innovation gains from CVC investments when espoused goals of the CVC program have
to do with strategic and not financial benefits, our findings clarify and contextualize their
claim. We suggest that positive technical change could be attributed primarily to one
98
dimension of the CVC-new venture fit: market development potential. Besides, we
provide evidence that espoused CVC goals do not necessarily match those in use: despite
the fact that the majority of the corporations claim to be strategically-oriented in their
CVC activities, over 50% of CVC investments are passive, and at best could only have
financial benefits for the corporation.
The study has certain limitations. In particular, it is restricted to a relatively short
four-year period which is best covered by multiple data sources. It may be that our
findings are specific to that particular wave of CVC activity and that some of our
conclusions could have been different otherwise. There could also be certain industry
heterogeneity. We attempt to address these issues by introducing the year and industry
dummies. Overall, the study extends the line of research started by Gompers, Lerner,
Dushnitsky, Lennox, Keil, Maula, Murray, Zahra, and others, clarifies and contextualizes
important findings reported by the CVC literature, and sheds light on some issues left
unanswered by the prior research.
99
References
Aernoudt, R. & San Jose, A. 2003. Executive Forum: Early Stage Finance and Corporate
Venture - Two Worlds Apart? Venture Capital, 5(4): 277-286.
Amit, R. & Schoemaker, P. J. H. 1993. Strategic Assets and Organizational Rent.

Strategic Management Journal, 14(1): 33-46.
Anokhin, S. & Schulze, W. 2006a. Corporate Venture Capital and Internal R&D: The
Anokhin, S. & Schulze, W. 2006b. Do CVC Children Make Better CVC Parents?
Working Paper.
Barney, J. 1991. Firm Resources and Sustained Competitive Advantage. Journal of

Management, 17(1): 99-120.

B6.
Bourgeois, L. J., III. 1981. On the Measurement of Organizational Slack. Academy of


Cheng, J. L. C. & Kesner, I. F. 1997. Organizational Slack and Response to

Environmental Shifts: The Impact of Resource Allocation Patterns. Journal of
Management, 23(1): 1-18.

Review, 80(3): 4-11.

Christopher, A. 2000. Corporate Venture Capital: Moving to the Head of the Class.
Venture Capital Journal, 40(11).
Cockburn, I., Henderson, R., & Stern, S. 2000. Untangling the Origins of Competitive
Advantage. Strategic Management Journal, 21(10/11): 1123-1145.

(Computer) Program.
100
Collinson, S. 2001. Knowledge Management Capabilities in R&D: A U.K.-Japan
Company Comparison. R&D Management, 31(3): 335-347.
Collis, D. J. 1994. Research Note: How Valuable are Organizational Capabilities?

Conner, K. R. 1991. A Historical Comparison of Resource-Based Theory and Five

Schools of Throught Within Industrial Organizational Economics: Do We Have a New
Theory of the Firm? Journal of Management, 17(1): 121-154.

Durand, R. & Vargas, V. 2003. Ownership, organization, and private firms' efficient use
of resources. Strategic Management Journal, 24(7): 667-675.



Policy, 34(5): 615-639.


EVCJ. 2002a. Corporate Venturers: The VC Industry's Fair-weather Friends? European

Venture Capital Journal(November): 18-23.
EVCJ. 2002b. Corporate Venturing Falls Off. European Venture Capital

Journal(October): 53.
Fare, R., Grosskopf, S., Norris, M., & Zhang, Z. 1994. Productivity Growth, Technical
Progress, and Efficiency Change in Industrialized Countries. American Economic
Review, 84: 66-83.
Fredriksen, O. & Klofsten, M. 2001. Venture Capitalists' Governance of Their Portfolio

Companies. Journal of Enterprising Culture, 9(2): 201-219.
101
Paper # 6725.

Quarter): 1-17.
Grant, R. M. 1991. The Resource-Based Theory of Competitive Advantage: Implications

for Strategy Formulation. California Management Review, 33(3): 114-135.
Hardymon, G. F., DeNino, M. J., & Salter, M. S. 1983. When Corporate Venture Capital
Doesn't Work. Harvard Business Review, 61(3): 114-120.

Economics, 64(2): 285-314.
Henderson, R. 1994. The Evolution of Integrative Capability: Innovation in

Cardiovascular Drug Discovery. Industrial and Corporate Change, 3(3): 607-630.
Henderson, R. & Cockburn, I. 1994. Measuring Competence? Exploring Firm Effects in

Pharmaceutical Research. Strategic Management Journal, 15(8): 63-84.
Hollingsworth, B. 2004. Non Parametric Efficiency Measurement. The Economic

Journal,, 114: F307-F311.
Kogut, B. & Zander, U. 1992. Knowledge of the Firm, Combinative Capabilities, and the
Replication of Technology. Organization Science: A Journal of the Institute of
Management Sciences, 3(3): 383-397.
Lawrence, P. R. & Lorsch, J. W. 1967. Organization and Environment: Managing

Differentiation and Integration. Homewood, IL: Irwin.
Lawson, M. B. 2001. In Praise of Slack: Time is of the Essence. Academy of

Management Executive, 15(3): 125-135.

23(3): 16-27.
Finance, 50(1): 301-318.
Malmquist, S. 1953. Index numbers and indifference surfaces. Trabajos de Estatistica, 4:

209-242.
102

Riyanto, Y. E. & Schwienbacher, A. 2005. The Strategic Use of Corporate Venture

Schilling, M. A. 2002. Technology Success and Failure in Winner-Take-All Markets:

The Impact of Learning Orientation, Timing, and Network Externalities. Academy of
Schilling, M. A., Vidal, P., Ployhart, R. E., & Marangoni, A. 2003. Learning by Doing
Something Else: Variation, Relatedness, and Learning Curve. Management Science,
49(1): 39-56.
Schoonhoven, C. B. 1981. Problems with Contingency Theory: Testing Assumptions

Hidden within the Language of Contingency "Theory". Administrative Science
Quarterly, 26(3): 349-377.
Shane, S. 2001. Technology Regimes and New Firm Formation. Management Science,
47(9): 1173-1190.
Shepherd, D. A. 1999. Venture capitalists' introspection: A comparison of "in use" and

"espoused" decision policies. Journal of Small Business Management, 37(2): 76-87.
Singh, J. V. 1986. Performance, Slack, and Risk Taking in Organizational Decision

Making. Academy of Management Journal, 29(3): 562-585.
Teece, D. J., Pisano, G., & Shuen, A. 1997. Dynamic Capabilities and Strategic
Management. Strategic Management Journal, 18(7): 509-533.

6(2): 341-377.
Utterback, J. M. 1974. Innovation in Industry and the Diffusion of Technology. Science,

183(February 15): 658-662.
103
Venkatraman, N. & Camillus, J. C. 1984. Exploring the Concept of "Fit" in Strategic
Management. Academy of Management Review, 9(3): 513-525.
Venkatraman, N. 1989. The Concept of Fit in Strategy Research: Toward Verbal and
Statistical Correspondence. Academy of Management Review, 14(3): 423-444.
Verona, G. 1999. A Resource-Based View of Product Development. Academy of

Wernerfelt, B. 1984. A Resource-based View of the Firm. Strategic Management Journal,

5: 171-180.
Winter, S. G. 2003. Understanding Dynamic Capabilities. Strategic Management Journal,

24(10): 991-995.
Zahra, S. A. 1996. Governance, Ownership, and Corporate Entrepreneurship: The

Moderating Impact of Industry Technological Opportunities. Academy of Management
Journal, 39(6): 1713-1735.
Zajac, E. J., Golden, B. R., & Shortell, S. M. 1991. New Organizational Forms for
Enhancing Innovation - the Case of Internal Corporate Joint Ventures. Management
Science, 37(2): 170-184.
104
Figure 3. Production Frontier Shift in Computer and Electronic Product Manufacturing Industry, 1998-2003
(colored dots represent companies, solid lines reflect the frontier)
105
Table 4. Descriptive Statistics
106
Table 5. Models Summary
107
108
Essay III. Do CVC Children Make Better CVC Parents?
ABSTRACT
Corporate venture capital (CVC) investments may bring a number of strategic
benefits to the incumbent corporations including the heavily sought-after ‘window on
technology’ or increase in corporate innovativeness. Yet, the number of corporations
actively pursuing CVC agenda remains modest. Using organizational imprinting and the
resource-based view of the firm as theoretical lenses this paper attempts to address two
questions: 1) what makes some corporations and not others to initiate CVC programs and
2) what makes some corporate parents better than others in absorbing the know-how
developed elsewhere.
The analysis demonstrates that history matters in structural decisions of the
corporations to establish CVC programs and in their ability to realize innovative potential
inherent in the new ventures’ ideas. Corporations started with the help of CVC are
significantly more likely to initiate CVC programs of their own. This is especially true
for CVC children parented by the corporations with clear policies of taking seats on the
boards of their CVC children. The results hold after controlling for a number of
organizational and industry variables. We also demonstrate that CVC children make
better CVC parents in terms of their ability to absorb the know-how developed elsewhere
which reflects in higher rates of realized corporate innovation. At the same time, they
demonstrate inferior ability to utilize their own research and development. These results
hold after controlling for organizational, industry, and temporal effects.
109
The study fits adequately with the previous work on CVC and sheds light on some
issues left unanswered by the prior research.
110
Do CVC Children Make Better CVC Parents?
1. Introduction
Corporations innovate (Schumpeter, 1942); yet successful innovation often
requires access to knowledge that is not available within organizational boundaries
(Cohen & Levinthal, 1990). To access this knowledge corporations design sophisticated
vehicles attempting to identify and capture external innovative ideas. One of such
vehicles that allows incumbents to tap into the know-how developed elsewhere
(Dushnitsky, 2004) is corporate venture capital (CVC) – equity investments in
entrepreneurial ventures by incumbent firms not made solely for financial reasons
(Chesbrough & Tucci, 2004; Dushnitsky & Lenox, 2005a).
Estimated two thirds of CVC programs are purely strategic in nature (Reaume,
2003); nine out of ten corporate venture capital programs list strategic objectives as at
least one of their main objectives (Hellmann, 2002). Of these, the majority look for
innovation, or ‘window on technology’ (Ernst & Young, 2002). Indeed, existing research
demonstrates that CVC investments are positively associated with both corporate
patenting (Dushnitsky et al., 2005a) and realized corporate innovation (Anokhin &
Schulze, 2006). Thus, to intensify corporate innovation, CVC investments may be
instrumental.
Yet, learning from CVC investments may require significant effort on the part of
the corporate parent. Dushnitsky et al. (2005a) show that harvesting innovations from
entrepreneurial ventures may be subject to significant industry effects and among other
things depends on the incumbent’s absorptive capacity. Anokhin et al. (2006)
111
demonstrate that realization of the effects of CVC investments is delayed: while internal
research and development affects corporate innovativeness almost immediately, CVC
effects, as a rule, become visible with a forward temporal lag of approximately two years.
Regardless of the potential benefits of CVC programs (innovation, new markets, etc.) the
share of the firms actively pursuing CVC agenda remains modest. One of the
explanations to this state of affairs is that managing a CVC program may require
specialized competencies and mechanisms which the majority of the corporations do not
readily possess and need to develop specifically. In this light, corporations that possess
such specialized capabilities to deal with external know-how are more likely to both
initiate CVC programs and quickly realize innovative potential of the new ventures’
ideas.
Literature on organizational imprinting and the resource based view of the firm
(RBV) help identify such corporations. Organizational imprinting framework maintains
that the firms would have a tendency to naturally perpetuate organizational arrangements
present at the moment of their inception. RBV urges corporations to preserve and employ
their unique competencies regardless of external environment pressures (Dierickx &
Cool, 1989). Taken together, these perspectives seem to suggest that corporations started
with the help of corporate venture capital (CVC children) may be more likely to establish
CVC programs of their own (become CVC parents). Such corporations also possess the
ultimate first-hand knowledge of the CVC process and are aware of potential pitfalls
unlike traditionally founded firms. As such, they are more likely to realize the effects of
their CVC investments relatively quickly.
112
In this study we propose to examine the effect of being a CVC child on becoming
a CVC parent. We also look at the speed with which innovative potential of the new
ventures is realized by corporate parents that once were CVC children and compare it
with how quickly traditionally founded corporations absorb the external know-how.
Organizational imprinting and the resource-based view of the firm form the basis for our
theoretical development. Empirical part builds on the data sources commonly used by the
CVC literature (e.g., Corporate Venturing Directory & Yearbook, Compustat, etc.).
However, we prefer to use the notion of realized innovation and not conventional patent-
related measures when looking at the rates of corporate innovation (Anokhin et al., 2006).
In so doing we attempt to avoid the problems typically associated with patents as proxy
for innovation; we are interested in the conversion of knowledge into economic benefits
(technical advancement) rather than codification of knowledge in patents.
The paper proceeds as follows. Theoretical arguments are introduced in Section 2.
Hypotheses development is summarized in Section 3. Section 4 explains methodological
aspects of the study. Results are presented in Section 5 and discussed in Section 6. The
paper concludes with the overview of the contributions of the study, its limitations, and
suggestions for future work.
2. Literature Review
2.1. Organizational Imprinting
While goals that corporations pursue when establishing CVC programs – such as
‘window on technology’, stimulating demand on own products and services, search for
acquisition targets, penetration to new markets, etc. – are sufficiently covered by the prior
113
research (Dushnitsky & Lenox, 2005b; Ernst & Young, 2002; Gompers, 2002; Riyanto &
Schwienbacher, 2001), why some corporations and not others make a commitment to
corporate venture investments (become CVC parents) is not well documented. All stated
goals provide strategic benefits to the corporations; yet, the share of the firms actively
pursuing CVC agenda remains modest. Organizational imprinting and the resource based
view of the firm provide unique insights into some of the factors that could explain the
organization’s decision to initiate a CVC program.
According to the organizational imprinting literature, establishing a CVC program
could be seen as one of the structural decisions and development trajectories that depend
to a large extent on the firm’s initial founding conditions. Firms have a tendency to
naturally perpetuate organizational arrangements present at the moment of their
inception: such arrangements are ‘imprinted’ in organizations and manifest themselves
throughout the firm’s tenure (Marquis, 2003; Stinchcombe, 1965). Initial conditions that
could affect the future trajectory of individual organizations include both the environment
at founding (Boeker, 1988; Meyer & Brown, 1977; Swaminathan, 1996) and the personal
characteristics of founders and top management (Baron, Hannan, & Burton, 1999;
Boeker, 1989; Eisenhardt & Schoonhoven, 1990).
History determines many aspects of the present structure of organizations
(Stinchcombe, 1965): early decisions and founding conditions imprint the firm, limit its
strategic choice, and continue to impact its long-term performance (Aspelund, Berg-
Utby, & Skjevdal, 2005; Bamford, Dean, & McDougall, 1999; Boeker, 1988, 1989).
Among aspects of the firm subject to imprinting are its overall structure and strategy; the
technology, patterns of activity, and routines that govern the organization’s day-to-day
114
activities; and organizational culture (Kriauciunas & Kale, 2002; Sastry & Coen, 2000).
Environmental conditions existing at the formation (founding) stage of an organization
that may imprint the firm and have a continued impact on the firm’s actions and
performance include technological, economic, legal, competitive, and social aspects
(Kriauciunas et al., 2002; Zyglidopoulos, 1999). For organizational processes
characterized by intense interaction with the outside parties initial conditions necessarily
include partner’s organizational routines, interface structure and partner’s expectations
(Doz, 1996).
According to Stinchcombe (1965), firm characteristics, as represented when
founded, often continue to be the best way to organize and do business. Even as the
organization matures through different stages of the life cycle, its characteristics are more
likely to reflect the requirements at the time of its founding than the current requirements
of the organization (Lynall, Golden, & Hillman, 2003). The early stages of a firm’s
existence see the development of the organization’s routines and cultures (such as the
policy on corporate venturing) that guide managerial decisions. In these initial stages, the
entrepreneurs must decide on an initial strategy by means of resources at hand and those
they can realistically acquire (Dollinger, 1999); the resources essential to the
organization’s purpose can only be obtained by the devices developed at that time
(Stinchcombe, 1965). Thus, if a particular organizational device appeared as prominent at
the time of the firm’s founding, it is likely to affect initial strategy of the new firm.
Furthermore, since the knowledge embodied in following a rule set, or undertaking a
task, is ‘remembered’ by an organization through routinization and becomes embedded in
115
structures and practices (Freel, 1998; Nelson & Winter, 1982), it is likely to retain
presence throughout organization’s development.
Organizational imprinting line of reasoning shares basic premises with the path
dependence arguments: today’s opportunities depend upon yesterday’s decisions (Lockett
& Thompson, 2001); once a particular course of action is committed to, only incremental
changes from that path are likely to result (Lynall et al., 2003). For instance, the
experience of innovating would develop capabilities within the firm that should
themselves assist further innovation and other nonroutine aspects of firm behavior
(Geroski, Machin, & Van Reenan, 1993). At the same time, imprinting framework per se
is more concerned with the effect of founding conditions on the social form or
organizational structure rather than a rule set or technological trajectory that the
organization follows. The mechanism by which initial conditions are imprinted in the
organization could be better understood if one examines a different level of analysis than
the main theory (Marquis, 2003; Stinchcombe, 1965). For community networks, one
needs to analyze the behavior of the component organizations; for imprinted
organizations – its individual members (Marquis, 2003).
New organizations generally involve new roles, which have to be learned. Former
occupants of roles – or, in the case of startups with no prior history, founders – teach their
successors not only skills but also decision criteria (Stinchcombe, 1965). It is in this
sense, that the organizations could be seen as a projection of the founders’ personae.
Interestingly, for the firms that are started with the help of CVC financing (CVC
children), incumbent corporations (corporate parents) assume certain aspects of both the
environment and the founder roles. Together with other environmental elements,
116
corporate parents define technological, economic, legal, competitive, and social aspects
of the new venture acumen (Kriauciunas et al., 2002; Zyglidopoulos, 1999). Corporate
parent’s organizational routines, interface structure and expectations contribute to the
initial conditions within which the new firm is founded. As founders (Gedajlovic,
Lubatkin, & Schulze, 2004), incumbent corporations are a source of at least some of the
firm’s initial equity capital, typically provide labor and technical expertise. Since
corporate parents play the roles of both environment and founder their imprinting
influence on the new firm is likely to be very significant. Norms subscribed to by the
corporate parents are likely to serve as especially important templates for newly
established firms and as such get perpetuated over time (Marquis, 2003).
2.2. The Resource-based View of the Firm
Organizational imprinting and path dependency literatures are rather
deterministic. In this sense, the resourced-based view (RBV) of the firm could be seen as
their natural extension (Barney, 1991; Conner, 1991; Dierickx et al., 1989; Wernerfelt,
1984). RBV is a useful perspective that explains how corporations build and sustain
competitive advantage. It attributes sustainable competitive advantage to superiority of
resources under the firm’s control. Resources include all assets, capabilities,
organizational processes, firm attributes, information, knowledge, etc. controlled by a
firm that enable the firm to conceive of and implement strategies that improve its
efficiency and effectiveness (Barney, 1991, p. 101, emphasis added). That is, a resource
does not have to be owned by the corporation to enable superior performance but has to
somehow be controlled.
117
This definition has an important implication for the study of corporate venture
capital. CVC investments guarantee a degree of control over the new ventures’ know-
how: the majority of the corporations take seats on the boards of the new ventures they
support, provide technical assistance to them, and perform the due diligence of their
business plans (Barry, 2000, 2001, 2002; Chesbrough, 2002). Thus, given careful
planning, organizing, and structuring, corporate venturing may become a useful resource
that corporations can use to conceive of and implement their strategies and develop into
an important capability in its own right (Barney, 1991; Porter, 1981).
Generally, capabilities could be defined as a high-level routine (or collection of
routines) that, together with its implementing input flows, confers upon an organization’s
management a set of decision options for producing significant outputs of a particular
type (Winter, 2003: p. 991). They refer to a firm’s capacity to deploy (a bundle of)
resources using organizational processes (Amit & Schoemaker, 1993; Grant, 1991).
Along with assets, organizational processes, firm attributes, information, knowledge, etc.,
capabilities are a part of the firm’s resource bundle (Barney, 1991).
Capabilities are always objective-specific; there are no “general purpose”
capabilities (Winter, 2003). Depending on the objective, literature differentiates between
technological, marketing (Verona, 1999), organizational (Collinson, 2001), integrative
(Henderson, 1994; Lawrence & Lorsch, 1967; Yeoh & Roth, 1999), combinative (Kogut
& Zander, 1992), dynamic capabilities (Teece, Pisano, & Shuen, 1997), dynamic
technical capabilities (Tripsas, 1997), etc. The latter (integrative, combinative, dynamic)
govern the rate of change of the former (which are referred to as ordinary or first-order)
118
capabilities and thus could be considered second-order capabilities (Collis, 1994; Winter,
2003). Similarly, capabilities could be considered second-order resources.
The capacity of the corporation to absorb and deploy externally developed know-
how accessed through CVC investments could be referred to as CVC capability. Properly
organized, it becomes valuable, rare, imperfectly imitable, and non-substitutable second-
order resource. This, according to Barney (1991), is the key to generate sustainable
superior outcomes. Dierickx et al. (1989) develop the argument further and suggest that
corporations should concentrate on their idiosyncratic skills and resources – such as CVC
capability – even at the expense of the competitive environment.
3. Hypotheses Development
As argued above, firms have a tendency to naturally perpetuate organizational
arrangements present at the moment of their inception. Initial environmental conditions
and founders’ characteristics affect future structural decisions of the new firms. Since in
the beginning CVC parents are both the environment and the founders for CVC children,
the exposure to the practices of corporate venturing that the new firms get is very
prominent. In fact, corporate parents actively establish their presence and instill their
values into startups: our calculations indicate that over 85 % of the corporations that have
a policy on taking seats on the new venture’s board do take advantage of this opportunity.
Besides, they provide technological, legal and organizational expertise to the new
ventures. Hence, the importance of corporate venture capital in the eyes of the new
venture is likely to be further magnified. That is, CVC programs are likely to be
imprinted in the CVC children as a legitimate and effective organizational arrangement.
119
Learning their roles from the corporate parent’s executives, new venture decision makers
are likely to perpetuate the concept of CVC and establish a program of their own once the
organization passes a certain stage in its life cycle. This should be particularly true when
CVC parents take seats on the new ventures’ boards. In other words, from the
organizational imprinting perspective companies started with the help of CVC, and
especially CVC children who had CVC parents’ representatives on their boards, are
expected to be more likely to start CVC programs of their own.
Besides, companies established with the help of corporate venture capital have the
most accurate first-hand experience with the corporate venturing processes. They know
the challenges of the CVC process and can cope with them; they are aware of the
reservations that the new ventures may have when dealing with corporate parents and
know how to address them. In short, through their path dependency, they possess the
CVC capability which by default is absent at most traditionally started corporations. In
line with the RBV and capabilities literatures, these companies are better off maintaining
and employing their CVC capability. Accordingly, CVC children can be expected to have
a higher propensity to start CVC programs of their own, that is, become CVC parents,
than corporations with no prior history of accepting CVC funding. Therefore:
Hypothesis 1. Corporations with prior history of accepting CVC financing are
more likely to initiate CVC programs than corporations without such history.
120
Hypothesis 2. CVC children with prior history of having CVC parents’
representatives on their boards are more likely to initiate CVC programs than
corporations without such history.
Since CVC children are more likely to possess well-developed CVC capabilities
than corporation with no prior history of accepting CVC funding, they should be able to
more quickly absorb the know-how of the supported venture and incorporate it into their
own activities. It has been shown that in general it takes considerable time to CVC
parents to realize the innovative potential of the new ventures (Anokhin et al., 2006). It
appears that CVC effects on the rates of corporate innovation are realized with the two
year forward temporal lag while own R&D effects become significant in the immediate
perspective. This time lag may be shortened if not eliminated for the corporations with
superior CVC capabilities.
Having been on the both sides of the CVC process, corporations established with
the help of CVC financing understand the intricacies of the issues facing new ventures
seeking corporate venture capital support. Such corporations should be able to provide
concerns that the new ventures might have. Accordingly, it seems plausible that the time
corporate innovation may be shorter for CVC children than for traditionally established
corporations. It follows that:
121
Hypothesis 3. Time required for the effects of CVC investments to be reflected in
the rates of realized corporate innovation is shorter for the corporations with prior
history of accepting CVC financing than for corporations without such history.
4. Data and Method
4.1. Data
To test our hypotheses we construct a unique dataset by matching multiple
secondary data sources. VentureXpert by Venture Economics and Corporate Venturing
Directory & Yearbook (hereafter – the Yearbook) by AssetAlternatives are utilized to
reconstruct the pattern of CVC investments by incumbent corporations, as well as to
identify CVC children and CVC parents1. VentureXpert contains detailed information on
the activities of the private equity industry and has been used in CVC research
extensively (Dushnitsky, 2004; Dushnitsky et al., 2005b, 2005a; Dushnitsky & Lenox,
forthcoming; Gompers & Lerner, 1998; Keil, Zahra, & Maula, 2004; Maula & Murray,
2000). The Yearbook covers a similar domain and has also been used previously by the
CVC literature (Birkinshaw & Hill, 2003; Chesbrough et al., 2004; Dushnitsky, 2002;
Gompers, 2002). Prior literature indicates that both these data sources have certain
deficiencies and may inflate the number of investment rounds (Lerner, 1994, 1995) or
double count particular deals (Anokhin et al., 2006). Besides, while sharing information
on many investments, each data source has information on some deals that are not
covered by the other database. Thus, by carefully matching the data we are able to obtain
the most accurate information on the CVC disbursements of the corporations.
1
122
VentureXpert covers the period from 1969 to the present time, while the Yearbook only
contains information on CVC investments committed during 1998-2001 (with few
exceptions). Accordingly, we operate with the data on investments done during this four-
year period.
Finally, we merge our database with the annual firm-level accounting and
financial data from Standard & Poor’s Compustat. Since the data reported in the
Compustat relate to a financial and not a calendar year of the corporation, we do not use
annual aggregates reported by VentureXpert directly but rather look at the exact dates of
particular deals to match them to appropriate financial years. Thus, for a corporation with
financial year starting in March and ending in February we would consider a CVC
investment made in January of 2000 as a part of year 1999. To test different time lags, we
look at the financial and accounting data over the six-year period of 1998-2003.
The merger of VentureXpert, the Yearbook, and Compustat yields a sample of
163 corporations that engaged in corporate venture capital investments during years 1998
to 20012. Finally, we augment the data set by matching these corporations to 187 similar
firms that did not participate in the CVC activity during the specified period. Additional
background check revealed that 44 of them have parented or co-parented CVC children at
some point in their lifetime. Thus, our final sample includes 350 corporations of which
207 could be classified as CVC parents. 65 out of 350 corporations had prior history of
accepting CVC financing at founding (were CVC children).
2
We deliberately excluded certain industries such as financial services, real estate, hotels, etc.
123
4.2. Variables and Estimation
Dependent variable for Hypotheses 1 and 2 is cvparent which takes a value of 1 if
the focal corporation has been recorded in either VentureXpert or the Yearbook as a CVC
investor or co-investor, and value of 0 otherwise. Dependent variable for Hypothesis 3 is
the rates of corporate innovation (RCI) which computes the ratio of technical efficiency
of the company at time t to technical efficiency at time (t-1). It reflects whether or not the
corporation has achieved technical advancement, that is found a way to produce more
output using a comparable resource bundle as a result of its CVC investments (Anokhin
et al., 2006; Penrose, 1959). Essentially, we make a comparison of sets of inputs and
outputs in different time periods for the corporations within their subindustries3.
Mathematically the algorithm is explained in Thursby & Thursby (2002). The
programming algorithm is realized in DEAP software (Coelli, 1996; Hollingsworth,
2004).
Key independent variable is cvchild which takes a value of 1 if the corporation
has prior experience of accepting CVC funding as recorded in VentureXpert or the
Yearbook and 0 otherwise. Another independent variable is cvboard which takes a value
of 1 if the CVC child has been parented or co-parented by a corporation that has a clear
policy of taking seats on their CVC children’s boards as reported in the Yearbook and 0
otherwise. We control for organizational slack (measured as the current ratio
(Chesbrough et al., 2004)), industry membership, and temporal effects (include dummy
variable for the year 2000 known for the crash of the dot.com market which constituted a
significant portion of the CVC investment portfolio). Since the likelihood of becoming a
3
Overall, 20 subindustries were analyzed
124
CVC parent may to some extent be an artifact of organizational size, in testing
Hypothesis 1 we control for it with the log of sales.
For Hypothesis 2 we also need to control for R&D intensity and CVC intensity as
both are important sources of corporate innovation (Anokhin et al., 2006). We calculate
R&D intensity as the ratio of R&D outlays to the dollar amount of the corporation’s
assets. Adjustment is made for the industry average as is commonplace in the strategic
management literature (Schilling, 2002). CVC intensity is measured as the number of new
ventures supported by the incumbent corporation in a given year. There are three main
reasons for choosing this measure over a more intuitive dollar amount one. First,
corporations tap into the knowledge developed by the new venture regardless of the size
of their investment (Reaume, 2003). Second, the amount invested is often simply a
function of the investment round: later rounds typically require more significant
investments (Gompers et al., 1998). Third, the Yearbook does not report amounts
invested by a particular corporation in individual deals and instead provides the overall
figure. Although our data allow us to estimate the dollar amount of those investments
such estimation may introduce unnecessary noise. In any case, our calculations indicate
that our measure of CVC intensity is highly correlated with the dollar amount of CVC
investments (r = 0.85). We test for the direct effects of cvchild on RCI and for the
interaction of cvchild with both R&D intensity and CVC intensity.
We employ logistic regression to test Hypotheses 1 and 2 (slack and size variables
are averaged over the four-year period), and feasible generalized least squares regression
with autocorrelation for Hypothesis 3. FGLS is required to address possible violation of
the independence assumption by the RCI variable (Thursby et al., 2002).
125
5. Results
Descriptive statistics can be found in Table 6. Overall, 59% of the corporations
included into the data base have parented or co-parented startups with the help of their
CVC programs. Over 18% of all corporations in the sample are CVC children, that is,
they accepted CVC financing at a founding stage. In case of corporations with CVC
programs active during 1998-2001 this number is over 25%, while for the firms not
active in corporate venture capital investments it is slightly below 13%. Over 35% of
CVC children have been parented or co-parented by the corporations that have clear
policies of taking seats on the boards of their CVC children. On average, corporations in
the sample experience technical advancement of about 1.09 in magnitude which indicates
rather substantial rates of realized corporate innovation. Corporations with active
corporate venture capital programs on average invest in 5.45 startups per year while the
most active companies may support over 250 new ventures (e.g., Intel in year 2000).
Results of hypotheses testing are presented in Table 7. They lend support to
Hypothesis 1 (Model 1). Overall, the model fits the data well (-2LL=419.44, Nagelkerke
R2=.17, Cox & Snell R2=.12) and correctly classifies 82% of the cases where the
corporation becomes a CVC parent. Corporations with prior history of accepting CVC
financing are significantly more likely to initiate CVC programs than corporations
without such history. Hypothesis 2 is also supported (Model 2). In addition to the positive
effect of accepting CVC financing at founding, CVC children parented or co-parented by
126
the corporations with a clear policy of taking seats on the boards of their CVC children
are significantly more likely to start CVC programs of their own than corporations
without such history. The model also demonstrates good fit (-2LL=415.64, Nagelkerke
R2=.18, Cox & Snell R2=.13). Hypothesis 3 is supported as well (Models 3-4). Even in
the long-term perspective (two-year lag) CVC children demonstrate (marginally)
significant advantage in realizing CVC effects over the corporations with no prior CVC
funding history (Model 3). They also show higher effectiveness in using CVC
investments to augment the rates of corporate innovation in the immediate perspective as
suggested by the positive effect of the interaction of cvchild and CVC intensity on RCI
(Model 4). At the same time, CVC children appear to demonstrate inferior results
converting their own R&D into realized innovation (Model 3). The findings are checked
with the Huber-White robust estimate of variance to account for overconfidence
potentially produced by FGLS as suggested by Wiggins (2001) (not reported here).
Overall, similar results are obtained.
6. Discussion
First, we would like to note the higher-than-expected proportion of CVC children
among analyzed corporations. Over 25% of the companies with active CVC programs
have prior history of accepting CVC financing at founding; the same is true of more than
12% of corporations not currently active in corporate venture capital investments. Taken
127
together, these numbers may imply that the phenomenon at question is substantially more
widespread than the extant research seems to indicate.
All our hypotheses find empirical support. Support for the imprinting-based
arguments is particularly strong. Controlling for other things, corporations with prior
history of accepting CVC financing are over 2 times more likely to initiate CVC
programs than corporations without such history, and almost 4 times more likely to do so
if they were parented or co-parented by a corporation with a clear policy of taking seats
on the boards of its CVC children. We suggest that this could be attributed to the
imprinting in such organizations of corporate venture capital as a legitimate and
successful means of accessing know-how developed elsewhere. The imprinting is likely
to happen because at the moment of these firms’ inception both initial environment and
founders – incarnated in the CVC parent – convey through a multitude of channels a
consistent message of corporate venturing being a potent mechanism of tapping into the
knowledge and ideas advanced by technological startups.
Control variables – organizational slack and organizational size – also turn out to
be significant as expected. Larger corporations with underutilized resources are more
likely to initiate corporate venture capital programs. Industry controls do not demonstrate
significance indicating that the process of becoming a CVC parent varies little across
different industries. This corroborates to some extent our conjecture that organizational
imprinting accounts for a substantial portion of the observed behavior. Namely, corporate
venturing is better described by the dynamics of social and not purely economic or
technological processes. It is imprinting of CVC as a valid instrument in the
organization’s toolkit (manifested, perhaps, in organizational norms, routines, decision
128
criteria, organizational culture) that makes some organizations and not others to seriously
consider initiating CVC programs in order to strategically screen and support new
technologies.
In other words, we see that path dependency applies in the context of corporate
venture capital. Surely, not all CVC parents can trace their heritage to being CVC
children and not all CVC children become CVC parents. At the same time, by tracing the
heritage of incumbent corporations to their founding days we see that firms that were
started with the help of corporate venture capital are likely to remain on this path. This
sheds light on why some corporations and not others initiate CVC programs.
CVC children are more likely to become CVC parents. We also see that they
make better CVC parents in terms of their ability to absorb the know-how of the new
venture which, all else equal, reflects in higher rates of realized corporate innovation.
CVC parents get more out of the new ventures they support with CVC money if they
themselves had prior experience of accepting CVC financing at founding. Not only do
they make better use of their CVC investments in the immediate perspective as indicated
by the positive significant interaction term in Model 4, but also their ability to learn from
CVC investments remains superior even when looked at with a two-year forward lag.
We also see that CVC intensity has a negative effect on the rates of realized
corporate innovation when considered in the immediate perspective. Overall, this finding
is consistent with Anokhin et al. (2006). We believe the reason here is that it takes some
time for the corporations to ‘naturalize’ essentially ‘foreign’ innovative ideas possessed
by the new ventures. It may also indicate that organizations need specialized skills to deal
with such foreign ideas. In this sense, the fact that CVC children are superior in their
129
ability to absorb the external know-how (as indicated by the positive significant
interaction term in Model 4) is perfectly in line with this conjecture. Earlier, we
suggested that CVC children have superior CVC capabilities. This finding seems to
provide some empirical support to this purely theoretical argument.
At the same time, we see that CVC children are far less likely to effectively utilize
their own R&D efforts. It could be argued that the corporations are forced to make a
choice between CVC and R&D in terms of where to invest their limited resources. Prior
literature indicates that such tensions may indeed exist (Anokhin et al., 2006). At the
same time, if this were also true in the case of CVC children, we should have seen
organizational slack as a significant covariate: the tensions would be less likely to appear
if the organization had significant amount of excess, unutilized or underutilized capacity.
Our analysis, however, does not indicate that organizational slack is significant in this
case (Models 3-4). Thus, we are inclined to suggest that it is not the economic
substitution of internal source of innovation with the external one that makes CVC
children less effective in realizing innovative potential of their internal R&D efforts.
Instead, what we see could probably be explained by the imprinted preferences of the
CVC children to use new ventures in search of innovative ideas, preferences that are
embodied in their routines, decision making rules and heuristics, and organizational
culture. In this light, it may be interesting to see whether or not accepting CVC funding at
founding impedes future ability of the CVC child to develop innovative projects
internally.
The fact that CVC children are better in realizing the innovative potential of their
CVC investments could also point to their superior CVC capabilities. Since they have
130
been on the both sides of the CVC process and understand the intricacies of the issues
facing new ventures seeking corporate venture capital support, they are able to provide
concerns that the new ventures might have. Accordingly, it is only natural that the time
corporate innovation is shorter for CVC children than for traditionally established
corporations.
7. Conclusion
In this study we demonstrate that path dependency applies in the context of
corporate venture capital. CVC children are more likely to become CVC parents. They
also make better CVC parents in terms of their ability to absorb the know-how of the new
venture which, all else equal, reflects in higher rates of realized corporate innovation
exhibited by such companies. Two possible explanations include organizational
imprinting and superior CVC capabilities possessed by the CVC children.
The study makes a number of contributions. First, it raises the question of why
some corporations and not others initiate CVC programs of their own. This adds to the
extant line of inquiry on corporate venture capital that so far has mainly focused on the
innovative implications of CVC investments for the corporate parent, benefits for the new
ventures of accepting CVC funding at founding, survival rates of entire CVC programs,
and other issues. Second, it adapts the arguments based on the organizational imprinting
framework to the new context – CVC financing – and thus extends the theoretical and
conceptual toolkit of the CVC research. Third, it demonstrates that socially driven
131
processes such as imprinting are not less important in understanding the dynamics of
corporate venturing than traditionally researched economic-related perspectives.
The study has certain limitations. In particular, it concentrates on CVC
investments committed during the third wave of CVC activity (late 1990s – early 2000s).
It may be that our findings are specific to that particular wave and that some of our
conclusions could have been different if we analyzed previous waves of CVC
investments as well (mid 1960s to early 1970s and late 1970s to late 1980s). After all,
each time period has its own dynamics, and it could be that a number of other reasons –
such as heavy preference of incumbent corporations to invest into internet-, and telecom-
related ventures – could have somehow distorted our results. We attempted to address
this issue by introducing the year and industry dummies although we acknowledge that
the way it could have affected our conclusions is unknown. Another matter to consider is
the possibility of the endogeneity problem. It may be that something not captured by our
models drives the likelihood of both becoming a CVC child and establishing a CVC
program. While we controlled for a number of things, and our results were robust to
different model specifications we would like to explore this issue further in our future
research.
Overall, the study extends the line of research started by Gompers, Lerner,
Dushnitsky, Lennox, Keil, Maula, Murray, Zahra, and others. It also suggests several
potential venues for future research. In particular, it may be interesting to see whether or
not accepting CVC funding at founding impedes future ability of the CVC child to
develop innovative projects internally. The study fits adequately with the previous work
132
on corporate venture capital and sheds light on some issues left unanswered by the prior
research.
133
References
Amit, R. & Schoemaker, P. J. H. 1993. Strategic Assets and Organizational Rent.

Anokhin, S. & Schulze, W. 2006. Corporate Venture Capital and Internal R&D: The
Aspelund, A., Berg-Utby, T., & Skjevdal, R. 2005. Initial Resource's Influence on New
Venture Survival: A Longitudinal Study of New Technology-based Firms. Technovation,
25: 1337-1347.
Bamford, C. E., Dean, T. J., & McDougall, P. P. 1999. An Examination of Founding

Conditions and Decisions Upon the Performance of New Bank Start-ups. Journal of
Business Venturing, 15: 253-277.
Barney, J. 1991. Firm Resources and Sustained Competitive Advantage. Journal of

Management, 17(1): 99-120.
Baron, J. N., Hannan, M. T., & Burton, D. 1999. Building the Iron Cage: Determinants of
Managerial Intensity in the Early Years of Organizations. American Sociological Review,
64: 527-547.

B6.
Boeker, W. 1988. Entrepreneurial and Environmental Imprinting at Time of Founding. In

G. R. Carroll (Ed.), Ecological Models of Organizations: 33-51. Cambridge, MA:
Ballinger.
Boeker, W. 1989. Strategic Change: The Effects of Founding and History. Academy of
Management Journal, 32: 489-515.

Review, 80(3): 4-11.
134

(Computer) Program.

Collinson, S. 2001. Knowledge Management Capabilities in R&D: A U.K.-Japan

Company Comparison. R&D Management, 31(3): 335-347.
Collis, D. J. 1994. Research Note: How Valuable are Organizational Capabilities?

Conner, K. R. 1991. A Historical Comparison of Resource-Based Theory and Five

Schools of Throught Within Industrial Organizational Economics: Do We Have a New
Theory of the Firm? Journal of Management, 17(1): 121-154.

Dollinger, M. J. 1999. Entrepreneurship: Strategies and Resources. Homewood, IL:

Richard D. Irwin.
Doz, Y. L. 1996. The Evolution of Cooperation in Strategic Alliances: Initial Conditions

or Learning Processes? Strategic Management Journal, 17(Summer): 55-83.



Policy, 34(5): 615-639.

Eisenhardt, K. M. & Schoonhoven, C. B. 1990. Organizational Growth: Linking

Founding Team, Strategy, Environment, and Growth among U.S. Semiconductor
Ventures, 1978-1988. Administrative Science Quarterly, 35: 504-529.
135
Freel, M. S. 1998. Evolution, Innovation and Learning: Evidence from Case Studies.
Entrepreneurship and Regional Development, 10: 137-149.
Gedajlovic, E., Lubatkin, M. H., & Schulze, W. 2004. Crossing the Threshold from
Founder Management to Professional Management: A Governance Perspective. Journal
of Management Studies, 41(5): 899-912.
Geroski, P., Machin, S., & Van Reenan, J. 1993. The Profitability of Innovating Firms.
RAND Journal of Economics, 24: 198-211.

Paper # 6725.

Quarter): 1-17.
Grant, R. M. 1991. The Resource-Based Theory of Competitive Advantage: Implications

for Strategy Formulation. California Management Review, 33(3): 114-135.

Economics, 64(2): 285-314.
Henderson, R. 1994. The Evolution of Integrative Capability: Innovation in

Cardiovascular Drug Discovery. Industrial and Corporate Change, 3(3): 607-630.
Hollingsworth, B. 2004. Non Parametric Efficiency Measurement. The Economic

Journal,, 114: F307-F311.
Kogut, B. & Zander, U. 1992. Knowledge of the Firm, Combinative Capabilities, and the
Replication of Technology. Organization Science: A Journal of the Institute of
Management Sciences, 3(3): 383-397.
Kriauciunas, A. & Kale, P. 2002. The Impact of Socialist Imprinting and Search for
Knowledge on Resource Change: An Empirical Study of Firms in Lithuania. William
Davidson Working Paper Number 446.
Lawrence, P. R. & Lorsch, J. W. 1967. Organization and Environment: Managing

Differentiation and Integration. Homewood, IL: Irwin.
136
23(3): 16-27.
Finance, 50(1): 301-318.
Lockett, A. & Thompson, S. 2001. The Resource-Based View and Economics. Journal of
Management, 27: 723-754.
Lynall, M. D., Golden, B. R., & Hillman, A. J. 2003. Board Composition from
Adolescence to Maturity: A Multitheoretic View. Academy of Management Review,
28(3): 416-431.
Marquis, C. 2003. The Pressure of the Past: Network Imprinting in Intercorporate

Communities. Administrative Science Quarterly, 48: 655-689.
Meyer, M. W. & Brown, M. C. 1977. The Process of Bureaucratization. American

Journal of Sociology, 83: 364-385.
Nelson, R. R. & Winter, S. G. 1982. An Evolutionary Theory of Economic Change.

Cambridge, MA: Harvard University Press.
Porter, M. 1981. The Contributions of Industrial Organization to Strategic Management.

Academy of Management Review, 6: 609-620.

Riyanto, Y. E. & Schwienbacher, A. 2001. On the Strategic Use of Corporate Venture

Sastry, A. & Coen, C. 2000. Beyond the Beginning: Building a Theory of Organizational
Imprinting. Working Paper.
Schilling, M. A. 2002. Technology Success and Failure in Winner-Take-All Markets:

The Impact of Learning Orientation, Timing, and Network Externalities. Academy of
Brothers.
137
Stinchcombe, A. L. 1965. Social Structure and Organizations. In J. G. March (Ed.),
Handbook of Organizations: 142-193. Chicago: Rand McNally.
Swaminathan, A. 1996. Environmental Conditions at Founding and Organizational

Mortality: A Trial-by-Fire Model. Academy of Management Journal, 39: 1350-1377.
Teece, D. J., Pisano, G., & Shuen, A. 1997. Dynamic Capabilities and Strategic
Management. Strategic Management Journal, 18(7): 509-533.

6(2): 341-377.
Verona, G. 1999. A Resource-Based View of Product Development. Academy of

Wernerfelt, B. 1984. A Resource-based View of the Firm. Strategic Management Journal,

5: 171-180.
Wiggins, V. 2001. Comparing xtgls with regress, cluster(). Stata FAQ resources, last
accessed on June 17, 2005 at http://www.stata.com/support/faqs/stat/xtgls_rob.html.
Winter, S. G. 2003. Understanding Dynamic Capabilities. Strategic Management Journal,

24(10): 991-995.
Zyglidopoulos, S. 1999. Initial Environmental Conditions and Technological Change.

Journal of Management Studies, 36(2): 241-262.
138
Table 6. Descriptive Statistics
139
Table 7. Regression Results
140

Case 1152821357

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Case 1152821357

Uploaded by

Copyright:

Available Formats

EMPIRICAL ESSAYS ON CORPORATE INNOVATION:

UNTANGLING THE EFFECTS OF CORPORATE VENTURE CAPITAL

Submitted in partial fulfillment of the requirements

for the degree of Doctor of Philosophy

Department of Marketing and Policy Studies

Weatherhead School of Management

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of

candidate for the Ph.D. degree *.

context of corporate innovation

Essay 3. Do CVC children make better CVC parents?

ESSAY 1. CORPORATE VENTURE CAPITAL AND INTERNAL R&D: THE INTERPLAY 24

IN THE CONTEXT OF CORPORATE INNOVATION

ESSAY 2. MAKING SENSE OF CORPORATE VENTURE CAPITAL II: TECHNOLOGY 68

PUSH VS. MARKET PULL

ESSAY 3. DO CVC CHILDREN MAKE BETTER CVC PARENTS? 109

Table 1. Descriptive Statistics 61

Table 2. Rates of Corporate Innovation and the CVC-R&D Interplay 62

Table 3. Industry Classification 66

Table 4. Descriptive Statistics 106

Table 5. Models Summary 107

Table 6. Descriptive Statistics 139

Table 7. Regression Results 140

Figure 1. Intertemporal Production Frontier in Computer and Electronic Product 60

Manufacturing Industry, 1998-2003

Figure 2. Interaction of CVC, R&D and Organizational Slack 63

Figure 3. Production Frontier Shift in Computer and Electronic Product 105

Manufacturing Industry, 1998-2003

dissertation. First and foremost, I would like to acknowledge my dissertation committee

me develop my interest in corporate venture capital.

contributing to my decision to pursue greater academic goals. I am especially grateful to

studies and to my children Veronica and Matthew for keeping me motivated.

and Jun Ye – for their friendship during the PhD program.

Untangling the Effects of Corporate Venture Capital

The dissertation combines three empirical essays on corporate venture capital

new ventures of accepting CVC funding at founding.

of the optimal size of CVC portfolio which maximizes corporate innovation.

the likelihood of fulfilling incumbents’ strategic objectives in different ways. Market

both rates of corporate innovation and scale efficiency decline regardless of

presence/absence of technological fit.

History matters in structural decisions of the corporations to establish CVC

own research and development.

Motivation for the Dissertation and Summary of the Appended Essays

Corporations are viewed as innovation engines (Schumpeter, 1942); yet

organizational boundaries (Cohen & Levinthal, 1990). To access this knowledge

innovative ideas. Corporate venture capital (CVC) – equity investments in

organizational arrangement allowing incumbents to tap into the know-how developed

At least three distinct sets of literatures have contributed to the development of

entrepreneurship (intrapreneurship). In particular, there is a literature that compares CVC

performance to that of independent and various forms of captive VC funds (government-

examines the innovative potential of different CVC program configurations (Dushnitsky

intrapreneurial development (Birkinshaw et al., 2003; Dushnitsky, 2002; Dushnitsky et

& Murray, 2000; Zahra, 1996; Zahra & George, 1999).

Corporate venturing research to date has developed an understanding of different

success rates of CVC programs (Chesbrough, 2002; Dushnitsky et al., 2005a,

literature on corporate venturing there is still no consensus as to whether and how

conceptual arguments are often contradictory or inconclusive. Researchers are still to

towards addressing these questions.

innovation and employs agency as a theoretical framework. It demonstrates that R&D

intensity is positively related to realized corporate innovation as measured by the rates of

intensity which is a particularly significant predictor of the rates of corporate innovation

The question of leadership in terms of the organizational hierarchy is also important:

detrimental for the success of the project.