Discussion Post Response

Is Open Source Software about Innovation? Collaborations with the Open Source Community and Innovation Performance of Software Entrepreneurial Venturesjsbm_356 340..364 by Evila Piva, Francesco Rentocchini, and Cristina Rossi-Lamastra

Practitioners generally assert that collaborations with the open source software (OSS) community enable software entrepreneurial ventures to achieve superior inno- vation performance. Nonetheless, scholars have never tested this assertion. This paper takes a first step toward filling this gap. First, based on the high-tech entrepreneurship literature and the OSS research stream, we illustrate why collaborations with the OSS community should exert a positive effect on entrepreneurial ventures’ innovation performance. Then, we provide a rigorous quantitative analysis of the innovation impact of these collaborations. Our econometric estimates indicate that entrepreneur- ial ventures collaborating with the OSS community exhibit superior innovation performance compared with their noncollaborating peers.

Introduction Conventional wisdom in the entrepre-

neurship field recognizes that entrepre- neurial ventures1 tend to have limited

financial and human capital to devote to research and development (R&D) activi- ties (Becker and Gordon 1966; Stevenson and Gumpert 1985). Therefore, these firms usually complement their internal

Evila Piva is assistant professor in the Department of Management, Economics and Industrial Engineering at Politecnico di Milano.

Francesco Rentocchini is research fellow in the Department of Economics of University of Trento and the Institute of Innovation and Knowledge Management (INGENIO) at Polytechnic University of Valencia.

Cristina Rossi-Lamastra is assistant professor in the Department of Management, Economics and Industrial Engineering at Politecnico di Milano.

Address correspondence to: Evila Piva, Department of Management, Economics and Indus- trial Engineering, Politecnico di Milano, via Lambruschini 4B, 20156 Milano. E-mail: evila.piva@ polimi.it. 1We define entrepreneurial ventures as new and independent (i.e., not controlled by third party organizations) firms established to commercialize novel ideas developed by their founders.

Journal of Small Business Management 2012 50(2), pp. 340–364

JOURNAL OF SMALL BUSINESS MANAGEMENT340

R&D efforts by networking intensively with external third parties (Stuart and Sorenson 2007) that have valuable knowledge and competences (“innova- tion inputs” hereafter).

Over the past decade, the astonishing developments in information and com- munication technologies (ICTs), along with the surge in new ICT-based methods of producing and circulating knowledge, have embedded entrepre- neurial ventures in an increasingly wide network of potential external collabora- tions. ICTs facilitate the access and in-sourcing of innovation inputs from distant individuals, firms, or universities (Piva, Grilli, and Rossi-Lamastra 2011).2

In addition, new prospective collabora- tors have emerged. Valuable innovation inputs are currently produced by com- munities of users and developers who interact through the Internet (Hargrave and Van de Ven 2006). In particular, intense and vibrant collaborations link the online community that develops open source software (“the OSS commu- nity” hereafter) and software entrepre- neurial ventures. Existing evidence suggests that many entrepreneurial ven- tures in the software industry belong to the category of firms that Bonaccorsi, Giannangeli, and Rossi (2006) has labeled OSS firms, indicating that they offer OSS-based software solutions to their customers (Dahlander 2007; Dahlander and Magnusson 2008). In this paper, we refer to these entrepre- neurial ventures as OSS entrepreneurial ventures.

OSS entrepreneurial ventures access and in-source innovation inputs pro- duced and distributed through the Inter- net by the OSS community, and use them to establish their offering. In other words, these firms download the OSS

code available on the Internet and develop new software solutions based on that code. These OSS solutions are tai- lored to specific customers’ needs or are released to the mass market under an OSS license. Alternatively, OSS entrepre- neurial ventures provide services (e.g., installation, system integration, or main- tenance) for well-established OSS products, such as Linux or Apache. More- over, OSS entrepreneurial ventures often participate in software development projects of the OSS community (Bonac- corsi and Rossi 2006), thereby actively contributing to the private provision of the OSS public good (Johnson 2002).

Collaborations between for-profit firms and the OSS community have been puzzling management and economic scholars since their inception. A lively research stream has examined the multi- faceted aspects of these collaborat- ions (for a review, see von Krogh, Rossi-Lamastra, and Haefliger 2012). Many studies have elucidated the peculiarities of OSS-based business models (Hecker 1999). Likewise, numerous contributions have focused on firms’ motivations to engage in the OSS movement (Rossi and Bonaccorsi 2005), on why and how firms usually mix the provision of proprietary and OSS solutions (see again Bonaccorsi, Giannangeli, and Rossi 2006), on the advantages and disadvantages of firms’ participation in OSS projects (Capra et al. 2011; Lerner, Pathak, and Tirole 2006), and on the organizational challenges posed by firm–community collaborations (Alexy and Leitner 2011; Colombo, Piva, and Rossi-Lamastra 2011; Dahlander and Wallin 2006). Two major gaps currently confront this literature. First, the real- world evidence documents that both large players in the software industry and small start-ups collaborate with the

2A long tradition in entrepreneurship research has recognized the importance of collaborations with firms and universities for entrepreneurial ventures’ innovation processes (for a recent review, see Hoskisson et al. 2011).

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 341

OSS community. Nevertheless, until now, few authors have taken into account the heterogeneity of OSS firms. Notable exceptions are Dahlander (2007) and Gruber and Henkel (2006). Both works have focused on new firms in the OSS domain. Dahlander (2007) has explored the different approaches to harness the OSS community adopted by de novo entrants, whereas Gruber and Henkel (2006) has investigated how the liabilities of newness and smallness of start-ups and market entry barriers affect new OSS firms. Despite these exceptions, entrepreneurship scholars have devoted scant attention to whether and how the distinctive characteristics of entrepre- neurial ventures shape their collabora- tions with the OSS community. Second, the relationship between OSS and inno- vation has been poorly investigated. Rossi-Lamastra (2009) has compared the innovativeness of OSS and proprietary solutions produced by small Italian firms collaborating with the OSS community. However, the author refers to the soft- ware solution and not to the firm as the unit of analysis. Indeed, limited research efforts have been applied to explore the effects of collaborations with the OSS community on firms’ innovation perfor- mance. An exception is Stam (2009), who has analyzed the effects of participation in OSS projects on firms’ innovation per- formance. However, the author has not compared the innovation performance of collaborating and noncollaborating firms.

This paper contributes to filling these gaps by theoretically and empirically addressing the following research ques- tion: Do entrepreneurial ventures that col- laborate with the OSS community (i.e., OSS entrepreneurial ventures) exhibit superior innovation performance in comparison with their noncollaborating peers?

Answering such a research question contributes to the general debate on the impact of collaborations with external third parties on firms’ innovation perfor-

mance, thus contributing to the academic discourse on open innovation. In addi- tion, by focusing explicitly on entrepre- neurial ventures, our analysis fits in the lively stream of the entrepreneurship lit- erature that examines how entrepreneur- ial ventures network with third parties for innovation purposes. Accordingly, we feel confident that our work will stimulate discourse among scholars in the fields of open innovation and entre- preneurship.

Moreover, our analysis has great prac- tical relevance. The allegedly positive effect that collaborations with the OSS community exert on the innovation per- formance of entrepreneurial ventures has strong echoes in the business and tech- nical press, and in conversations among practitioners. Survey data document the positive view that entrepreneurs seem to have regarding the relationship between collaborations with the OSS community and firm innovation performance. The ELISS I survey of Italian software firms (for details, see Bonaccorsi and Rossi 2004) analyzes the motivations driving firms’ collaborations with the OSS com- munity. The top-ranking motive selected by the 146 respondents was that collabo- rating with the OSS community allows even new and small firms to be innova- tive (Rossi and Bonaccorsi 2005). This result was confirmed by a second wave of the survey (ELISS II; for details, see Bonaccorsi, Rossi, and Scateni 2005) of approximately 900 European software firms. In addition, case study evidence has indicated that collaborations with the OSS community enhance the innovation performance of entrepreneurial ventures because resources that are freely avail- able within the OSS community can be used as low-cost inputs for firms’ inno- vation processes (Dahlander and Mag- nusson 2005, 2008). Nonetheless, to the best of our knowledge, it has never been tested whether collaborations with the OSS community promote innovation by entrepreneurial ventures.

JOURNAL OF SMALL BUSINESS MANAGEMENT342

In the empirical part of this paper, the impact of collaborations with the OSS community on the innovation perfor- mance of entrepreneurial ventures is rig- orously analyzed through the estimation of econometric models. The empirical analysis takes advantage of a unique data set that contains detailed information on collaborations with the OSS community and the innovation activity of 199 Italian entrepreneurial ventures observed during the period 2005–2008.

The paper is organized as follows. In the next section, we extensively illustrate why one might expect collaborations with the OSS community to exert a posi- tive effect on the innovation perfor- mance of entrepreneurial ventures. In the Methods section, we illustrate the data set, describe the sample used in the empirical analysis, specify the economet- ric models, and describe the variables included in the models. The Results section summarizes the results of the econometric estimates. The last section synthesizes the main findings, acknowl- edges the limitations of the study, and indicates directions for further research.

Conceptual Background Entrepreneurial ventures in the soft-

ware industry (“software entrepreneurial ventures” hereafter) are usually at a dis- advantage in the innovation race. Like most entrepreneurial ventures operating in high-tech industries (“high-tech entre- preneurial ventures” hereafter), software entrepreneurial ventures generally lack financial resources (Carpenter and Petersen 2002a, 2002b), internal compe- tences (Colombo and Piva 2008), and complementary assets (Teece 1986).

Rooting on the high-tech entrepreneur- ship literature and the OSS research stream, we argue that collaborations with the OSS community render the earlier mentioned obstacles to innovation (as mentioned earlier) less severe. Accord- ingly, entrepreneurial ventures that col- laborate with the OSS community (i.e.,

OSS entrepreneurial ventures) exhibit superior innovation performance in com- parison with their noncollaborating peers.

Collaborations with the OSS Community and Software Entrepreneurial Ventures’ Financial Constraints

High-tech entrepreneurial ventures usually have limited internal finances (Carpenter and Petersen 2002a) and poor access to debt financing (Carpenter and Petersen 2002b). Moreover, empirical work has shown that there is a substan- tial wedge between the costs of internal and external equity financing (see, e.g., Asquith and Mullins 1986). This situation inhibits access to equity capital for most new high-tech ventures, especially in countries with a less developed and/or bank-based financial system (Berger and Udell 1998). Software entrepreneurial ventures are no exception. Existing evi- dence suggests that these ventures gen- erally have few financial resources to invest in innovation activities (e.g., Romijn and Albaladejo 2002). Collabora- tions with the OSS community reduce the negative impact of these financial constraints of OSS entrepreneurial ven- tures. As a consequence, OSS entrepre- neurial ventures enjoy an advantage in innovation over their noncollaborating peers.

By developing valuable OSS solu- tions, OSS entrepreneurial ventures may attract external capital more easily. The relationship between collaborations with the OSS community and a firm’s ability to attract external financing has gone generally unnoted in the academic lit- erature (for two exceptions, see Alexy 2008; Feller and Fitzgerald 2002). However, the professional press notes that collaborations with the OSS com- munity usually impress venture capita- lists favorably, thereby making them more willing to sponsor the innovation projects of OSS entrepreneurial ven- tures. OSS has, indeed, been regarded

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 343

as reliable and of high quality since the entrance of major players in the soft- ware industry, such as IBM and Sun, into the OSS arena. In addition, the openness of the OSS code allows anyone to inspect it to assess its value (Lerner and Tirole 2005), thereby signal- ing good quality for the OSS entrepre- neurial ventures that have contributed to its development. It might be claimed that OSS entrepreneurial ventures also have a disadvantage regarding the attraction of external financing. Venture capitalists may, indeed, find it unattrac- tive to invest in firms producing techno- logical artifacts for which it is impossible to enforce intellectual prop- erty rights. Scholars have documented that venture capitalists attach a value to patent holding when deciding to finance a software firm (Mann and Sager 2007). Nevertheless, to the best of our knowledge, no study has shown that the possible fear of openness negatively counterbalances venture capi- talists’ positive assessment regarding OSS.

Two additional compelling reasons argue in favor of a positive impact of collaborations with the OSS community on OSS entrepreneurial ventures’ finan- cial constraints. First, and probably most important, collaborations with the OSS community substitute for internal R&D activities. The OSS community consti- tutes a common pool of software code and programming competences that OSS entrepreneurial ventures can access at a cost that, in most cases, is very low. OSS entrepreneurial ventures can then use these code and competences as inputs to develop new products and services (Bonaccorsi and Rossi 2003, 2004). Second, the free availability of inputs from the OSS community reduces the costs of OSS entrepreneurial ventures’ daily operations. For example, no license fees are required to use an OSS compiler when developing software (Lerner and Tirole 2005). Likewise, user-to-user assis-

tance with mailing lists maintained by OSS developers (Lakhani and von Hippel 2003) reduces the costs of OSS entrepre- neurial ventures’ customer care. To sum- marize, collaborations with the OSS community generate positive pecuniary externalities (Antonelli 1995). Therefore, OSS entrepreneurial ventures can invest the freed financial resources in their innovation processes.

Collaborations with the OSS Community and Software Entrepreneurial Ventures’ Lack of Internal Competences

Initially, the competences of high- tech entrepreneurial ventures largely coincide with those of their founders (Cooper and Bruno 1977; Feeser and Willard 1990). However, these initial competences may be insufficient for the development of a sustained stream of innovations (McMullen and Shepherd 2006). Therefore, high-tech entrepre- neurial ventures soon experience a compelling need to enlarge their initial competence endowment. Competence enlargement might be achieved by hiring new talented individuals (Baron 2010; Baron and Hannan 2002). However, scholars agree that high-tech entrepreneurial ventures usually have difficulties in recruiting (and subse- quently retaining) talented individuals (for a review of the studies on this topic, see Colombo and Rossi-Lamastra 2011). Moreover, these firms rarely have the resources to make major invest- ments in personnel (Bryant and Allen 2009). Therefore, high-tech entrepre- neurial ventures usually expand their internal competence by partnering with other firms or collaborating with univer- sities (for a recent review on this theme, see Hoskisson et al. 2011). In this framework, collaborations with the OSS community may be an alternative valuable competence-enlargement strat- egy for software entrepreneurial ven- tures for several reasons.

JOURNAL OF SMALL BUSINESS MANAGEMENT344

First, the OSS community is a large pool of talented individuals from which OSS entrepreneurial ventures can scout brilliant programmers, thereby comple- menting their internal competences with fresh individual skills (Eilhard 2008; Henkel 2009). OSS entrepreneurial ven- tures can access programmers’ human capital through different employment modes, for example, by hiring program- mers or by contracting them as free- lancers (Lepak and Snell 1999). In both cases, the OSS framework attenuates the problems that high-tech entrepre- neurial ventures usually experience in accessing human resources (Connelly et al. 2011).

However, it might be claimed that sin- gling out the OSS programmers whose skills match the firm’s requirements and linking these individuals to an OSS entre- preneurial venture is not so simple. Indeed, because anyone can enter the OSS community, OSS programmers have highly variable skills and capabilities (Colombo, Piva, and Rossi-Lamastra 2011). Nevertheless, the characteristics of OSS code might simplify this task. According to the OSS definition, OSS code must not only be open, but it must also contain a declaration of authorship.3

Consequently, an OSS entrepreneurial venture that is searching for talented pro- grammers has the opportunity to down- load the source code of a piece of OSS from the web, inspect it to assess its quality, trace the programmer who wrote it, and evaluate whether it is worthwhile to establish a collaboration with the pro- grammer. We do not want to deny that this might be a complex and time- consuming procedure; what we intend to say is that OSS entrepreneurial ventures can rely on a potentially powerful and valuable instrument for scouting talented

programmers that cannot be used by their noncollaborating peers.

Second, OSS entrepreneurial ventures can easily expand their internal compe- tences by interacting with other OSS firms within the community. The OSS community is being increasingly popu- lated by firms (Fitzgerald 2006), and evi- dence suggests that OSS firms are keen on interacting among each other and sharing their competences. OSS practitio- ners generally assert that OSS is more about collaboration than about competi- tion. This coopetition tendency (Van de Vrande et al. 2009) in the OSS realm is demonstrated by the numerous collabo- rative networks linking OSS firms that are developing software solutions based on the same OSS platform. The Japspor- tal network, which connects OSS firms working with the OSS platform Japs,4

and the network of OSS firms basing their business on the Zope content management system5 are prominent examples. In addition, academic research finds that OSS entrepreneurial ventures tend to inherit the values of knowledge sharing and collaboration of their funders who frequently have served as OSS developers (Dahlander 2007).

To summarize, OSS entrepreneurial ventures enjoy an advantage in the inno- vation race over their noncollaborating peers in that they can access an alterna- tive and, in many respects, valuable pool of external competences.

Collaborations with the OSS Community and Software Entrepreneurial Ventures’ Lack of Complementary Assets

High-tech entrepreneurial ventures frequently lack the complementary assets that they need to profit from inno- vation (Teece 1986). In principle, these

3http://www.opensource.org/osd.html. 4http://www.entando.com/portal/. 5http://www.zope.com.

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 345

firms could develop such complementary assets internally or gain access to them by establishing alliances with third parties. However, both of these solutions are hardly feasible for high-tech entre- preneurial ventures. The well-known lack of internal resources restrains high- tech entrepreneurial ventures from inter- nally developing complementary assets. Likewise, alliances often require high- tech entrepreneurial ventures to incur relevant transaction and management costs (Colombo, Grilli, and Piva 2006). These costs usually have a hindering effect on alliance formation. OSS entre- preneurial ventures can overcome these obstacles by using the many external complementary assets available from the OSS community (Dahlander and Wallin 2006). Indeed, OSS entrepreneurial ven- tures can choose applications that complement their innovative focal solu- tions from the OSS common pool of soft- ware programs and modules, instead of developing them from scratch or licens- ing them from other firms. No license fee is required when firms use these comple- mentary applications, with the compli- ance that the OSS licenses under which these applications are released is the only restriction to their use (McGowan 2001). It is worth noting that being com- pliant with OSS licenses is crucial for OSS entrepreneurial ventures aiming at integrating OSS complementary applica- tions into their offering. Indeed, achiev- ing such compliance requires an awareness of the varied legal provisions associated with the diverse OSS licenses (Lerner and Tirole 2005). Gaining such knowledge may be particularly time- consuming (Dahlander and Magnusson 2008), whereas failing to comply with the provisions of open source licenses can engender negative consequences on OSS entrepreneurial ventures. For example, if a complementary application is released under a copy-left license (e.g., the General Public License) and its integration with a software solution

developed by an OSS entrepreneurial venture requires the modification of the source code of the two programs, the OSS entrepreneurial venture has to release under a copy-left license the entire resulting software code (including the program that it has produced inter- nally). Indeed, copy-left licenses contain an inheritance provision that forbids the release under a proprietary license work that contains even one line of code taken from a copy-left program (Rosen 2001). The enforceability of OSS licenses has rarely been tested in the courts (Lerner and Tirole 2002). However, to establish effective collaborations with the OSS community, OSS entrepreneurial ven- tures must be trusted by community par- ticipants. If OSS developers envision that an OSS entrepreneurial venture may hijack OSS code and make it proprietary because it is not aware of the OSS license provision or, even worse, because it wants to make a profit on it, OSS programmers will be reluctant to provide feedback and contributions to the firm.

Finally, the OSS community is a low- cost channel for distributing and market- ing software programs (West and O’Mahony 2008). OSS entrepreneurial ventures can take advantage of the OSS distribution infrastructure based on online software repositories and dedi- cated websites, which enable OSS firms to reach a larger customer base at lower cost. To summarize, OSS entrepreneurial ventures enjoy an advantage in the inno- vation race over their noncollaborating peers because they can access the complementary assets made available to the OSS community.

Methods The Sample

The sample used in this paper was extracted from an original database developed in 2009 by the general admin- istration of the Emilia-Romagna region of Italy within the Emilia-Romagna Open

JOURNAL OF SMALL BUSINESS MANAGEMENT346

Source Survey (EROSS survey) project.6

The EROSS survey was intended to collect information about the collabora- tions between the OSS community and the software entrepreneurial ventures located in the region by administering a structured questionnaire to the owner– managers of these firms. The questions concerned information on firms’ struc- tural characteristics, sales and employee evolution, OSS offerings, strategies for external knowledge sourcing and intel- lectual property right (IPR) protection, and innovation processes. The data refer to the 2005–2008 period.

The design of the survey and the construction of the database underwent a careful preparation phase. At the beginning of 2009, in-depth face-to-face interviews were conducted with the owner–managers of three OSS entrepre- neurial ventures. All of the informants were interviewed once and were asked about the main themes that the EROSS research group intended to include in the questionnaire. The interviews lasted for 45–60 minutes and were conducted by two people, with one researcher posing the questions and the other taking notes. The results of this prelimi- nary analysis were used to design the questionnaire.

EROSS researchers intended to admin- ister the questionnaire to a representa- tive sample of software entrepreneurial ventures located in the Emilia-Romagna region. For this purpose, the regional population of software entrepreneurial ventures was identified by selecting the industry segments that include these firms from the Italian Classification of Economic Activities ATECO 2002. The resulting population included 7,355 entrepreneurial ventures. Next, a subset of 512 target firms was extracted. This subset was stratified according to the

province (the third level in the Nomen- clature of Territorial Units for Statistics codes) of firm location and industry segment. Between October and Decem- ber 2009, the owner–managers of the 512 target entrepreneurial ventures were contacted, and 297 were available for a telephone interview (response rate: 58 percent) based on the questionnaire described earlier. Before conducting the telephone interviews, the questions that might be more subject to selective memory problems were sent by e-mail or fax to the respondents who were asked to search for this information in advance. Therefore, even though there might be a recall bias in the data used in this paper, its extent is likely to be relatively limited. The fact that the data we use in our study are not subjective and can be verified by respondents also makes it unlikely that our results are driven by a common- method bias.

Because data regarding entrepreneur- ial ventures’ innovation performance were collected during the 2005–2008 period, to avoid reverse causality pro- blems, our sample does not include the entrepreneurial ventures established in 2005 or after, or firms that began collaborating with the OSS community in 2005 or after. Therefore, the sample used in this paper includes 199 firms. This sample is representative of the regional population of the 7,355 software entrepreneurial ventures by province, industry segment, and firm age (c2[7] = 5.5, c2[3] = 3.64, and c2[3] = 4.17, respectively).

Of the 199 sample firms, 30.3 percent (i.e., 60 entrepreneurial ventures) col- laborated with the OSS community, and thus were considered OSS entrepreneur- ial ventures. This figure is in line with the results of other surveys (e.g., Bonaccorsi, Giannangeli, and Rossi 2006).

6See http://www.regionedigitale.net/projects-piter/research-and-development for further details.

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 347

Table 1 reports some descriptive sta- tistics for the 199 sample firms and dis- tinguishes the 60 OSS entrepreneurial ventures from their 139 noncollaborating peers. Most sample firms were small; on December 31, 2005, 85.4 percent of them had 10 or fewer employees. On the same date, 71.4 percent of the sample firms had been operating for more than five years. Interestingly, the figures in the table indicate that the OSS entrepreneur- ial ventures in the sample are smaller and newer than their noncollaborating peers. However, c2 tests show that there are statistically significant differences between the distribution of OSS entre- preneurial ventures across age categories and the corresponding distributions of noncollaborating sample firms (c2[1] = 3.95), whereas there are no sig- nificant differences across size categories (c2[1] = 1.44).

At the survey date, the sample firms offered products and services in different product and service categories. We con- sidered the following four product cat- egories: (1) management applications, (2) office automation products, (3) web products (including content management systems, websites, portals, hosting, and e-commerce solutions), and (4) other types of products. Regarding services, we considered six categories: (1) instal- lation, (2) maintenance and assistance, (3) training, (4) integration, (5) software customization, and (6) other types of ser- vices. The vast majority of the sample firms did not offer products in more than one category but offered services in at least two categories. Most sample firms offered web products (81 firms, 40.7 percent of the sample), whereas few firms offered office automation products (51 firms, 25.6 percent). Regarding ser- vices, 73.9 percent of the sample firms offered maintenance and assistance ser- vices, whereas the least common service, with the exclusion of the residual cat- egory described as other types of ser- vices, was integration (60 firms, 30.2

percent). Interestingly, both the product portfolio and the service portfolio of the entrepreneurial ventures collaborating with the OSS community seem, on average, to be more diversified than those of their noncollaborating peers. Indeed, the percentage of OSS entrepre- neurial ventures in each product (service) category is greater than the cor- responding percentage for the entrepre- neurial ventures not collaborating with the OSS community. In this study, c2 tests indicate that in all of the categories, the differences between the percentages of OSS entrepreneurial ventures and the corresponding percentages of noncol- laborating firms are significant, the only exceptions being the product category management applications, and the two service categories of software customiza- tion and other types of services.

Innovation Performance of the Sample Firms

This section provides empirical evi- dence for the innovation performance of the sample entrepreneurial ventures. In this and the following sections, we focus only on innovation in terms of the introduction of new (or significantly improved) software solutions. To collect data on this type of innovation, the respondents of the EROSS survey were asked (1) whether their firms introduced any new or significantly improved soft- ware solutions between 2005 and 2008, (2) how many new or significantly improved software solutions their firms introduced between 2005 and 2008, and (3) what share of their firms’ total turn- over between 2005 and 2008 was from new or significantly improved software solutions.

These questions are similar to typical questions included in innovation surveys in general and in the CIS Community Innovation Survey in particular (for a review on innovation surveys, see Mairesse and Mohnen 2010). Therefore, we use these questions to draw three

JOURNAL OF SMALL BUSINESS MANAGEMENT348

T a b le

1 D

e sc

ri p ti

v e

S ta

ti st

ic s

o f

th e

S a m

p le

F ir

m s

S a m

p le

F ir

m s

(N =

1 9 9 )

E n tr

e p re

n e u ri

a l

V e n tu

re s

C o ll a b o ra

ti n g

w it

h th

e O

S S

C o m

m u n it

y (N

= 6 0 )

E n tr

e p re

n e u ri

a l

V e n tu

re s

N o t

C o ll a b o ra

ti n g

w it

h th

e O

S S

C o m

m u n it

y (N

= 1 3 9 )

N o .

P e rc

e n t

N o .

P e rc

e n t

N o .

P e rc

e n t

E m

p lo

y e e s

o n

D e ce

m b e r

3 1 ,

2 0 0 5

� 1 0

1 7 0

8 5 .4

5 4

9 0 .0

1 1 6

8 3 .4

>1 0

2 9

1 4 .6

6 1 0 .0

2 3

1 6 .5

T o ta

l 1 9 9

1 0 0 .0

6 0

1 0 0 .0

1 3 9

1 0 0 .0

A g e

cl as

s o n

D e ce

m b e r

3 1 ,

2 0 0 5

� 5

Y e ar

s 5 7

2 8 .6

2 3

3 8 .3

3 4

2 4 .5

>5 Y

e ar

s 1 4 2

7 1 .4

3 7

6 1 .7

1 0 5

7 5 .5

T o ta

l 1 9 9

1 0 0 .0

6 0

1 0 0 .0

1 3 9

1 0 0 .0

P ro

d u ct

C at

e g o ri

e s

M an

ag e m

e n t

A p p li ca

ti o n s

7 4

3 7 .2

2 7

4 5 .0

4 7

3 3 .8

O ffi

ce A

u to

m at

io n

P ro

d u ct

s 5 1

2 5 .6

2 3

3 8 .3

2 8

2 0 .1

W e b

P ro

d u ct

s 8 1

4 0 .7

3 8

6 3 .3

4 3

3 0 .9

O th

e r

T y p e s

o f

P ro

d u ct

s 5 7

2 8 .6

2 6

4 3 .3

3 1

2 2 .3

S e rv

ic e

C at

e g o ri

e s

In st

al la

ti o n

1 0 1

5 0 .8

3 8

6 3 .3

6 3

4 5 .3

M ai

n te

n an

ce an

d A

ss is

ta n ce

1 4 7

7 3 .9

5 1

8 5 .0

9 6

6 9 .1

T ra

in in

g 6 9

3 4 .7

2 7

4 5 .0

4 2

3 0 .2

In te

g ra

ti o n

6 0

3 0 .2

3 0

5 0 .0

3 0

2 1 .6

S o ft w

ar e

C u st

o m

iz at

io n

1 0 7

5 3 .8

3 7

6 1 .7

7 0

5 0 .4

O th

e r

T y p e s

o f

S e rv

ic e s

1 3

6 .5

4 6 .7

9 6 .5

O S S ,

O p e n

S o u rc

e S o ft w

ar e .

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 349

measures of entrepreneurial ventures’ innovation performance.

Using the answers to the first ques- tion, we built DInnovation, a dummy variable equaling 1 if the focal entrepre- neurial venture introduced any new soft- ware solutions during the 2005–2008 period. DInnovation is an indicator of the probability that an entrepreneurial venture innovates. The answers to the second question were used to build NIn- novation, a count variable equal to the number of new software solutions intro- duced by the focal entrepreneurial venture during the 2005–2008 period winsorized at 95 percent. NInnovation quantifies the innovations introduced by an entrepreneurial venture. This measure is similar to those used in highly cited papers (e.g., Acs, Audretsch, and Feldman 1992; Audretsch and Feldman 1996; for an application in the OSS realm, see Stam 2009). Using the third question, we built ShareInnSales, an

ordered variable that takes a value of 0 if the share of the focal firm’s turnover from new software solutions introduced during the 2005–2008 period equals 0 percent, 1 if it ranges between 0 percent and 20 percent, and 2 if it is higher than 20 percent. ShareInnSales captures the capacity of the innovations introduced by the focal entrepreneurial venture to create value in the medium term for the firm. In addition, the share of a firm’s turnover from new products is an estab- lished measure of the firm’s innovation (e.g., Laursen and Salter 2006).

Table 2 and Figure 1 report descrip- tive statistics for the three measures of innovation performance, distinguishing OSS from noncollaborating entrepre- neurial ventures. The table and figure reveal that the sample OSS entrepreneur- ial ventures exhibited superior innova- tion performance during the 2005–2008 period compared with their noncollabo- rating peers. Specifically, Table 2 indi-

Table 2 Innovation Performance of the Sample Firms between

2005 and 2008

Entrepreneurial Ventures Collaborating

with the OSS Community (N = 60)

Entrepreneurial Ventures Not Collaborating

with the OSS Community (N = 139)

No. Percent No. Percent

DInnovation =0 26 43.3 98 70.5 =1 34 56.7 41 29.5 Total 60 100.0 139 100.0

ShareInnSales =0 (i.e., 0 percent) 15 28.3 66 52.4 =1 (i.e., 1–20 percent) 11 20.8 31 24.6 =2 (i.e., >20 percent) 27 50.9 29 23.0 Total 53 100.0 126 100.0

OSS, Open Source Software.

JOURNAL OF SMALL BUSINESS MANAGEMENT350

cates that the OSS entrepreneurial ventures exhibited a greater probability of developing new software solutions; indeed, 56.7 percent of the OSS entrepre- neurial ventures introduced new or sig- nificantly improved software solutions between 2005 and 2008 versus 29.5 percent of the noncollaborating entrepre- neurial ventures (c2[1] = 13.17). The table also shows that the OSS entrepreneurial ventures achieved greater shares of turn- over from new software solutions than did their noncollaborating peers. In the former group, the share of turnover from software solutions introduced during the 2005–2008 period was greater than 20

percent for most firms (50.9 percent), whereas in the latter group, the share was greater for only 23 percent of the firms. Moreover, the share was equal to 0 percent for 28.3 percent of the OSS entrepreneurial ventures, whereas it equaled 0 percent for most noncollabo- rating firms (52.4 percent). The distribu- tion of the two groups of firms is clearly significantly different (c2[2] = 14.32).

The histogram in Figure 1 shows the distribution of the number of new or significantly improved software solu- tions introduced between 2005 and 2008 for both OSS entrepreneurial ven- tures and their noncollaborating coun-

Figure 1 Distribution of the Number of New or Significantly

Improved Solutions Introduced by Sample Firms between 2005 and 2008

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 351

terparts. For both groups of firms, the distribution is highly skewed, and the median number of new software solu- tions equals 0. However, the mean number is higher for the OSS entrepre- neurial ventures (6.6 versus 2.2), and a Wilcoxon rank-sum test indicates that this difference is significant at a 1 percent level.

To determine whether these differ- ences remain true once one controls for other factors that might influence the innovation performance of software entrepreneurial ventures, we performed the econometric analysis described in the following section.

Specification of the Econometric Models and Description of the Independent Variables

We analyzed the innovation impact of collaborations with the OSS community through the estimation of the following model:

Y DOSSCollaboration Z

i i T

i i

= + + +

α β γ ε , (1)

where Yi is one of the three measures of innovation performance presented in the previous section (i.e., DInnovation, NIn- novation, and ShareInnSales), DOSSCol- laborationi is the variable that identifies OSS entrepreneurial ventures, Zi is a vector of firm-specific control variables, and ei is the error term.

Overall, five different econometric models were estimated, according to the dependent variable under consideration. We ran a logit model when the depen- dent variable was the dummy DInnova- tion, Poisson and negative binomial7

models when we measured innovation through the count variable NInnovation,8

and an ordered logit model when the dependent variable was the ordered vari- able ShareInnSales.

The independent variables of the models described earlier are as follows. The key explanatory variable is DOSSCol- laboration, which is a dummy variable equaling 1 if the focal entrepreneurial venture started collaborating with the OSS community before 2005. If the coefficients of DOSSCollaboration were positive, we could conclude that entre- preneurial ventures collaborating with the OSS community exhibit innovation performance superior to that of their noncollaborating peers.

The models also included a series of control variables that might affect the innovation performance of software entrepreneurial ventures. Control vari- ables refer to 2005, and they are, there- fore, predetermined with respect to our dependent variables. First, we controlled for an entrepreneurial venture’s human capital (HumanCapital). In line with the literature on labor and education eco- nomics (Folloni and Vittadini 2010; Heckman, Lochner, and Todd 2003), we use a measure of human capital pertain-

7We adopted the negative binomial specification to control for departure from the assumption of equidispersion that is made when the model is of a Poisson type (i.e., E[NInnovation] = Var[NInnovation]). 8To ensure the reliability of our results regarding the impact of collaborations with the community on NInnovation, we adopted two alternative specifications. First, we used a weighted quasi-maximum likelihood Poisson model. Second, because NInnovation has a non-negligible number of zeros, we estimated both Zero-Inflated Poisson and Hurdle Poisson models (for a review of the solutions proposed in the econometric literature to have consistent estimates when a dependent variable is characterized by an excess of zero observations, see Cameron and Trivedi 2005). The results do not differ from those discussed later. These results are available from the authors upon request.

JOURNAL OF SMALL BUSINESS MANAGEMENT352

ing to off-the-job training. In particular, HumanCapital is equal to the percentage of the individuals employed by the focal firm that possessed at least a master’s- level education.9 Next, we included a proxy for firms’ R&D investments: Devel- opersShare. This variable is the number of software developers divided by the total number of employees of the focal firm. This measure is a widely used proxy for R&D investments in studies focusing on the software industry (see, e.g., Bessen and Hunt 2007), and proves to be espe- cially suitable for entrepreneurial ven- tures for which it is difficult to collect data on R&D expenses. Moreover, because it is widely accepted that firm size is likely to affect firm innovation performance (for a thorough survey, see Becheikh, Landry, and Amara 2006), we included LnSize, measured as the logarithm of the total number of employees (plus one), as a control.

To control for the sample firms’ prob- ability to introduce innovations other than new (improved) software solutions, we included a dummy equaling 1 if the focal entrepreneurial venture introduced any process or organizational innovation during the 2005–2008 period (DOtherIn- novation). To control for the use of formal instruments for the protection of intellectual property, we inserted in the model DIPR a dummy variable that equals 1 if the firm had ever used patents and/or trademarks. Because recent

studies have shown that access to exter- nal sources of information positively affects firms’ innovation potential (see, e.g., Laursen and Salter 2006), we con- trolled for the total number of external sources of information to which the entrepreneurial venture had access and could, thus, exploit for innovation pur- poses (NInfSources). Six potential sources of information were considered: (1) suppliers, (2) customers, (3) competi- tors, (4) public research organizations, (5) professional associations and online communities, and (6) social networks. Therefore, NInfSources is a cardinal vari- able ranging from 0 to 6. To control for the effect of agglomeration economies, we included a geographical dummy vari- able equaling 1 for the entrepreneurial ventures located in the Province of Bologna, Italy. Indeed, of the nine prov- inces of the Emilia-Romagna region, Bologna is the area in which most soft- ware entrepreneurial ventures are located (more than 28 percent of the entire population). Finally, we controlled for firm age and for market-specific effects by including a series of age dum- mies10 and industry segment dummies.11

Table 3 illustrates descriptive statistics for the independent variables and reports their correlation matrix. In general, the correlation across the inde- pendent variables is low, suggesting the absence of any relevant problems of multicollinearity.

9As an alternative proxy for the human capital of the entrepreneurial venture, we built a continuous variable calculated as the total number of years of schooling of employees. When we replace HumanCapital with this new indicator, the sign and significance of the coefficients of DOSSCollaboration do not change. The results are available from the authors upon request. 10We considered four age dummies that, respectively, equal 1 if the studied firm was founded (1) in 2004 (2) in 2003 (3) in 2002, and (4) in 2001. The baseline is being founded before 2001. 11We considered four product market segments and six service types. The product market segments are the following: (1) management applications; (2) software for office automation; (3) content management systems, websites, portals, hosting, e-commerce solutions; and (4) other products. The six service types are the following: (1) installation, (2) maintenance and support, (3) training, (4) integration of different components, (5) software customization, and (6) other services.

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 353

Results The Impact of Collaborations with the OSS Community on Firm Innovation Performance

Table 4 presents the estimates of the econometric models. For each model, specification (1) includes only the con- trols, whereas specification (2) also includes the key explanatory variable DOSSCollaboration. The first two columns of the table report the results of the logit models, with DInnovation being the dependent variable (Models 1a,b). The next four columns report the results, with NInnovation as the depen- dent variable differentiating between the Poisson model (Models 2a,b) and the negative binomial model (Models 3a,b). The last two columns report the results for the ordered logit models, with the share of sales of new software solutions, ShareInnSales, as the dependent variable (Models 4a,b).

Regarding the control variables (see Model specifications a), the coefficient of DOtherInnovation is positive and signifi- cant at conventional confidence levels in all of the specifications. Reasonably enough, a complementarity exists among the different innovation typologies (this result is in line with Cassiman and Veugelers 2006). Interestingly, all of the remaining determinants of the three measures of innovation performance are different. In Model 1a, none of the remaining controls is significant. Models 2a and 3a indicate that the number of new software solutions introduced by entrepreneurial ventures in the period under scrutiny was higher the larger the firm size was; LnSize, indeed, exhibits a positive coefficient, which is significant at a 1 percent level. Moreover, some dif- ferences exist between the Poisson and the negative binomial specifications regarding the significance of the con- trols. In the Poisson Model 2a, Human- Capital exhibits a positive coefficient that is significant at a 1 percent level.

T a b le

3 D

e sc

ri p ti

v e

S ta

ti st

ic s

a n d

C o rr

e la

ti o n

M a tr

ix o f

th e

E x p la

n a to

ry V

a ri

a b le

s (N

= 1 9 9 )

V a ri

a b le

M e a n

S ta

n d a rd

D e v ia

ti o n

M in

M a x

(1 )

(2 )

(3 )

(4 )

(5 )

(6 )

(1 )

D O

SS C

o ll

a b o ra

ti o n

0 .3

0 0 .4

6 0 .0

0 1 .0

0 1 .0

0 (2

) H

u m

a n

C a

p it

a l

4 .4

1 1 4 .3

1 0 .0

0 1 0 0 .0

0 0 .1

5 1 .0

0 (3

) D

ev el

o p er

sS h

a re

6 6 .9

0 3 6 .0

3 0 .0

0 1 0 0 .0

0 0 .1

3 0 .0

8 1 .0

0 (4

) L n S iz

e 1 .5

6 0 .8

8 0 .6

9 5 .9

3 -0

.0 4

0 .1

-0 .2

2 1 .0

0 (5

) D

O th

er In

n o va

ti o n

0 .3

0 0 .4

6 0 .0

0 1 .0

0 0 .1

2 -0

.1 1

-0 .0

4 0 .1

2 1 .0

0 (6

) D

IP R

0 .1

6 0 .3

6 0 .0

0 1 .0

0 0 .1

1 0 .0

4 -0

.1 2

0 .1

9 0 .1

5 1 .0

0 (7

) N

In fS

o u

rc es

1 .9

9 1 .3

4 0 .0

0 6 .0

0 0 .2

0 0 .0

5 0 .0

8 -0

.0 1

0 .3

0 0 .1

4

JOURNAL OF SMALL BUSINESS MANAGEMENT354

T a b le

4 R

e su

lt s

o f

th e

E co

n o m

e tr

ic E st

im a te

s o f

th e

E ff

e ct

o f

C o ll a b o ra

ti o n s

w it

h th

e O

p e n

S o u rc

e S o ft

w a re

C o m

m u n it

y o n

F ir

m In

n o v a ti

o n

P e rf

o rm

a n ce

D In

n o v a

ti o n

N In

n o v a

ti o n

S h

a re

In n

S a

le s

L o g it

P o is

so n

N e g a ti

v e

B in

o m

ia l

O rd

e re

d L o g it

(M o d e l

1 a )

(M o d e l

1 b )

(M o d e l

2 a )

(M o d e l

2 b )

(M o d e l

3 a )

(M o d e l

3 b )

(M o d e l

4 a )

(M o d e l

4 b )

a 0

C o n

st a

n t

-3 .2

4 **

* -3

.4 5 **

* -4

.1 4 **

* -3

.7 7 **

* -5

.7 2 **

* -5

.5 8 **

* –

– (1

.0 0 )

(1 .0

6 )

(0 .9

7 )

(0 .9

7 )

(0 .9

9 )

(1 .0

2 )

a 1

D O

SS C

o ll

a b o ra

ti o n

– 0 .9

9 **

– 0 .8

4 **

* –

0 .8

7 **

– 0 .8

1 **

(0 .4

0 )

(0 .2

9 )

(0 .4

3 )

(0 .4

0 )

a 2

D ev

el o p er

Sh a

re 0 .0

1 0 .0

1 0 .0

1 0 .0

0 0 .0

2 **

* 0 .0

2 **

* 0 .0

1 *

0 .0

1 *

(0 .0

1 )

(0 .0

1 )

(0 .0

0 )

(0 .0

0 )

(0 .0

1 )

(0 .0

1 )

(0 .0

1 )

(0 .0

1 )

a 3

H u

m a

n C

a p it

a l

0 .0

1 0 .0

0 0 .0

3 **

* 0 .0

3 **

* 0 .0

2 0 .0

1 -0

.0 1

-0 .0

1 (0

.0 1 )

(0 .0

1 )

(0 .0

1 )

(0 .0

1 )

(0 .0

1 )

(0 .0

1 )

(0 .0

1 )

(0 .0

1 )

a 4

L n

Si z e

0 .3

4 0 .3

4 0 .8

7 **

* 0 .8

3 **

* 1 .0

7 **

* 1 .0

6 **

* 0 .3

7 *

0 .3

8 *

(0 .2

1 )

(0 .2

1 )

(0 .1

8 )

(0 .1

7 )

(0 .2

3 )

(0 .2

3 )

(0 .2

1 )

(0 .2

1 )

a 5

D O

th er

In n

o va

ti o n

0 .8

2 **

0 .7

5 *

0 .8

8 **

0 .8

2 **

1 .2

2 **

* 1 .4

3 **

* 0 .8

7 **

0 .8

5 **

(0 .4

1 )

(0 .4

2 )

(0 .3

7 )

(0 .3

8 )

(0 .4

1 )

(0 .4

4 )

(0 .3

6 )

(0 .3

6 )

a 6

D IP

R 0 .0

6 0 .0

2 0 .2

7 0 .2

3 0 .1

0 -0

.2 1

0 .5

2 0 .4

3 (0

.4 5 )

(0 .4

4 )

(0 .4

6 )

(0 .4

1 )

(0 .4

2 )

(0 .4

6 )

(0 .5

2 )

(0 .5

0 )

a 7

N In

fS o u

rc es

-0 .0

5 -0

.0 8

0 .1

1 0 .0

6 0 .0

2 -0

.0 5

0 .0

8 0 .0

6 (0

.1 6 )

(0 .1

5 )

(0 .1

5 )

(0 .1

6 )

(0 .1

5 )

(0 .1

5 )

(0 .1

5 )

(0 .1

4 )

A g e

D u m

m ie

s In

c. In

c. In

c. In

c. In

c. In

c. In

c. In

c. G

e o g ra

p h ic

al D

u m

m ie

s In

c. In

c. In

c. In

c. In

c. In

c. In

c. In

c. In

d u st

ry D

u m

m ie

s In

c. In

c. In

c. In

c. In

c. In

c. In

c. In

c. T e st

s o f

P ro

p o rt

io n al

it y

o f

O d d s

ac ro

ss R e sp

o n se

C at

e g o ri

e s

B ra

n T e st

– –

– –

– –

– -0

.8 3

(1 2 )

A p p ro

x im

at e

L ik

e li h o o d

R at

io T e st

– –

– –

– –

– 2 8 .9

8 (2

1 )

M cF

ad d e n ’s

R 2

0 .1

6 0 .1

8 0 .4

8 0 .5

1 0 .0

8 0 .0

9 0 .1

6 0 .1

7 O

b se

rv at

io n s

1 9 9

1 9 9

1 9 9

1 9 9

1 9 9

1 9 9

1 7 9

1 7 9

W al

d c2

3 4 .5

7 (2

1 )*

* 3 6 .6

6 (2

2 )*

* 1 3 5 .1

1 (2

1 )*

** 1 4 9 .5

3 (2

2 )*

** 1 2 7 .1

5 (2

1 )*

** 1 2 0 .5

8 (2

2 )*

** 6 4 .0

8 (2

1 )*

** 6 3 .0

2 (2

2 )*

** L o g -L

ik e li h o o d

-1 1 1 .2

9 -1

0 8 .2

9 -7

4 2 .7

1 -7

0 8 .5

2 -2

8 5 .3

6 -2

8 4 .2

3 -1

6 0 .2

2 -1

5 8 .0

2

*p <

.1 0

** p

< .0

5 **

*p <

.0 1

R o b u st

st an

d ar

d e rr

o rs

an d

d e g re

e s

o f

fr e e d o m

ar e

in p ar

e n th

e se

s.

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 355

This indicates that the entrepreneurial ventures having superior human capital introduced more new software solutions between 2005 and 2008. Conversely, in Model 3a, DeveloperShare exhibits a positive and significant (at 1 percent) coefficient. Reasonably enough, the entrepreneurial ventures employing a higher share of software developers introduced a higher number of new soft- ware solutions. Finally, in Model 4a, both DeveloperShare and LnSize exhibit posi- tive and (weakly) significant coefficients.

When DOSSCollaboration is added to the set of regressors (see Models b), the signs and significance of the coeffi- cients of the controls do not differ from those in Models a. The insertion of DOSSCollaboration improves the explanatory power of the models, as is documented by the increase of the McFadden’s R2.12 More interestingly, the coefficient of DOSSCollaboration is posi- tive and significant in all Models b. This finding indicates that during the period under consideration, OSS entrepreneur- ial ventures had a higher probability of generating innovations, introduced a higher number of innovative solutions, and had higher sales from new software solutions than their noncollaborating peers. Overall, we find evidence that OSS entrepreneurial ventures exhibit innova- tion performance superior to that of the other software entrepreneurial ventures.

To assess the magnitude of the effects of collaborations with the OSS commu- nity on the innovation performance of software entrepreneurial ventures, based on the estimates of Models b, we first computed the three innovation measures for a “benchmark” entrepreneurial venture (i.e., with all the dummy and

cardinal control variables set at their median value and all the continuous control variables set at their mean value) with DOSSCollaboration = 0. Then, we calculated the estimated values of the three innovation measures for the same benchmark firm with DOSSCollabora- tion = 1. Regarding the benchmark entre- preneurial venture noncollaborating with the OSS community, the OSS entrepre- neurial venture was found to have a 23 percent higher probability to innovate, to introduce two additional new software solutions, and to have a 13 percent greater probability of being in a higher category of ShareInnSales.

It is worth noting that the estimates discussed earlier are unbiased and con- sistent, provided that the assumption of exogeneity of DOSSCollaboration is sat- isfied. However, some unobserved factors might render the observed rela- tionship between DOSSCollaboration and the measures of innovation perfor- mance endogenous. For example, both the decision to engage in collaborations with the OSS community and entrepre- neurial ventures’ innovation performance may be positively correlated with the demographic characteristics of entrepre- neurial ventures’ owner–managers (Beckman and Burton 2011; regarding the role played by owner–managers in shaping the innovation performance of high-tech entrepreneurial ventures, see also the recent contribution of Knockaert et al. 2011). Specifically, more experi- enced and educated owner–managers may be more likely to search for new knowledge sources, and thus to initiate collaborations with the OSS community. At the same time, conventional wisdom in entrepreneurship suggests that owner–

12We also run likelihood-ratio tests for the exclusion of additional variables from the restricted logit model and Lagrange multiplier tests of the generalized logit model. Both types of test are strongly rejected at a 1 percent significance level. This means that Model 1b is more informative and that linearity in the parameters can be confidently assumed. The results of the tests are available from the authors upon request.

JOURNAL OF SMALL BUSINESS MANAGEMENT356

managers with superior human capital are likely to positively influence the inno- vation performance of their firms.

Addressing the Endogeneity Problem

Controlling for the potential endoge- neity of DOSSCollaboration may be tricky because all of the models pre- sented and discussed in the previous section are nonlinear models, and DOS- SCollaboration is a dummy variable. Indeed, for continuous outcome vari- ables, two-stage regression strategies have been developed to tackle the problem, even when the endogenous variable is a dummy (Heckman 1978). However, in the presence of noncontinu- ous outcomes with endogenous dummy variables, an approach via maximum likelihood is needed to obtain a consis- tent and efficient estimator.

For this purpose, we relied on a series of endogenous switching regression models (Miranda and Rabe-Hesketh 2006). In particular, we resorted to the following specification13:

Y DOSSCollaboration Z

i i T

i i

= + + +

α β γ ε

1 1

1 (2)

DOSSCollaboration DOpenStandard

DOpenValues Z

i

i

i T

i

= + + +

α β β γ

2 2

3 ++ ε2i (3)

where Yi is one of the three measures of firm innovation performance (i.e., DInnovation, NInnovation, or Share-

InnSales). Equation (2) contains the same explanatory variable and controls included in equation (1) (see previous section). Equation (3) specifies a model for the endogenous switching dummy DOSSCollaboration. The model specifies two exclusion restrictions in the form of two dummies equaling 1 if the respon- dent for firm i rated as a highly important motive for collaborating with the OSS community sharing the OSS values (DOpenValues) and the possibility of exploiting the benefits on an open stan- dard (DOpenStandard), respectively.14

DOpenValues and DOpenStandard are good instruments because they are likely to explain the probability of collaborat- ing with the OSS community but not the innovation performance of the entrepre- neurial venture. The system of equations is fitted by maximum likelihood with Gauss–Hermite quadrature.

Moreover, to check for the presence of endogeneity, we tested the null hypothesis that the correlation between the error terms of the two equations (e1i and e2i) does not significantly differ from zero. When the dependent variables are DInnovation and ShareInnSales, the likelihood-ratio test of the correlation of the residuals of equations (2) and (3) is not rejected at conventional confidence levels (c2[1] = 0.21 and c2[1] = 1.72, respectively). Conversely, when we measure innovation performance through the count variable NInnovation, the likelihood-ratio test is rejected at a 1 percent significance level. Therefore, in this case, DOSSCollaboration is likely to

13Technical details on the econometric models, as well as a hands-on tutorial on how to estimate them on Stata, are provided in Miranda and Rabe-Hesketh (2006). 14We asked the respondents to rate on a four-point Likert scale their level of agreement with the following statements: (1) one of the key motives for collaborating with the OSS community is sharing the OSS values, and (2) one of the key motives for collaborating with the OSS community is the possibility of exploiting the benefits of an open standard. The scale ranged from 1 (strongly agree) to 4 (strongly disagree). Next, we assigned a value of 1 to DOpenValuesi and DOpenStandardi if the respondents rated their level of agreement with statements (1) and (2), respectively, as 1 or 2.

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 357

be endogenous. Nevertheless, the sign and significance of the coefficient of DOSSCollaboration are still in line with the results presented in Table 4.

Discussion and Conclusions

In this paper, we analyze the effects of collaborations with the OSS community on the innovation performance of soft- ware entrepreneurial ventures. By taking advantage of a unique data set and through the application of rigorous econometric techniques, we show that OSS entrepreneurial ventures exhibit superior innovation performance in com- parison with their noncollaborating peers. This finding is robust to the use of different innovation measures.

The paper contributes to the previous knowledge in three main respects. First, it adds to the extensive academic debate on open innovation. To date, the com- munity of scholars studying open innovation has mainly focused on col- laborations of established firms with other companies or universities (Laursen and Salter 2006). Entrepreneurial ven- tures being able to activate open innova- tion processes has been recognized only recently by the open innovation litera- ture (van de Vrande et al. 2009). More- over, the case of OSS is popular in the open innovation realm (van de Vrande, Vanhaverbeke, and Gassmann 2010) to the extent that the principles of OSS development have been used as a basis for understanding the functioning of open innovation (West and Gallagher 2006). However, how entrepreneurial ventures operate in this peculiar open innovation realm is still a poorly explored issue (for an exception, see Gruber and Henkel 2006).

Second, in the spirit of the present special issue, our work adds to the stream of the entrepreneurship literature that analyzes the impact of collabora- tions with external third parties on the innovation performance of entrepreneur-

ial ventures. A well-established tradition in the entrepreneurship field documents that entrepreneurial ventures intensively network with external third parties for innovation purposes (Stuart and Soren- son 2007). This literature has extensively studied the innovation impact of alli- ances with other firms and collaborations with public research organizations (e.g., Lee, Lee, and Pennings 2001). Con- versely, in general, no remarks have been made on the impact of collabora- tions with communities of users and developers. By documenting that net- working with the OSS community posi- tively affects entrepreneurial ventures’ innovation performance, this paper sug- gests that alternative third parties (i.e., unconventional allies, O’Mahony and Bechky 2008) can be sources of advan- tages in the innovation race for entrepre- neurial ventures.

Third, the present paper contributes to the OSS literature. The business and technical press has often speculated on the links between OSS and innovation, asking whether OSS is about innovation or merely imitates successful proprietary solutions. However, few academic studies have explicitly addressed innova- tion in the OSS realm (for recent excep- tions, see Katsamakas and Georgantzas 2010; Rossi-Lamastra 2009). In addition, no studies have addressed how collabo- rations with the OSS community affect the innovation performance of entrepre- neurial ventures. Stam (2009) has ana- lyzed the effects of participation in OSS projects on firms’ innovation perfor- mance. However, the author has neither explicitly considered the specificities of entrepreneurial ventures nor compared the innovation performance of collabo- rating and noncollaborating firms.

The paper has several limitations that suggest avenues for future research. First, we distinguish entrepreneurial ven- tures that collaborate with the OSS com- munity (OSS entrepreneurial ventures) from their noncollaborating peers by

JOURNAL OF SMALL BUSINESS MANAGEMENT358

introducing a dummy variable in our econometric specifications. However, an in-depth analysis of entrepreneurial ven- tures’ collaborations with the OSS com- munity would require more nuanced constructs and measures. OSS entrepre- neurial ventures are heterogeneous in their collaboration modes. Frequently, they mix the offering of proprietary and OSS solutions, thus adopting a hybrid business model (Bonaccorsi, Giannan- geli, and Rossi 2006; Lerner and Schankerman 2010). Moreover, different firms usually take advantage of the inno- vation inputs in-sourced from the OSS community in different ways, which range from selling prepackaged OSS products to developing new OSS program from scratch. These diverse ways of relying on the OSS community for innovation purposes may differently affect OSS entrepreneurial ventures’ innovation performances. In addition, OSS entrepreneurial ventures are possi- bly heterogeneous in their inner charac- teristics. These ventures probably have diverse internal competences and resources, which are likely to moderate the impact of collaborations with the OSS community on their innovation per- formance. Specifically, the role of human capital in determining the innovation performance of high-tech entrepreneur- ial ventures is well established in the field of entrepreneurship (for a review, see Unger et al. 2011). Therefore, it might be conjectured that the human capital of entrepreneurial ventures’ owner–managers and employees affects the relationship between collaborations with the OSS community and innovation. Owner–managers and employees with advanced education in computer science or with extensive experience in the OSS field are probably more able to appropri- ate the innovation benefits stemming from collaborations with the OSS com- munity. Specifically, experience in col- laborating with the OSS community is likely to play a major role. Leveraging

the OSS community for innovation pur- poses is far from simple for an OSS firm. Indeed, the OSS community is a varied common pool of resources and compe- tences, the quality of which is highly variable (Colombo, Piva, and Rossi- Lamastra 2011). In order to in-source valuable innovation inputs from such a common pool, an OSS entrepreneurial venture must learn to navigate it (Piva, Rentocchini, and Rossi-Lamastra 2011). The firm must develop familiarity with which OSS projects develop high-quality code to be used as a basis for developing new software solutions and with who are the most talented OSS developers to become involved with in innovation activities. Likewise, the OSS literature has shown that OSS developers are sometimes suspicious of firms because they fear that firms’ profit-oriented motives can drive them to hijack the open code, thereby removing it from the OSS common pool (O’Mahony 2003). Accordingly, OSS entrepreneurial ven- tures aiming at leveraging the OSS com- munity for innovation must learn how to cope with the values of openness and knowledge sharing that inform the OSS community. Violation of these values can severely undermine fruitful collabora- tions with the OSS community.

Second, in line with a mainstream tradition in the innovation field, this paper focuses on product innovation. However, scholars explicitly recognize that collaborating with the OSS commu- nity is an important organizational inno- vation in firms’ software production (Lerner and Tirole 2002). To exploit its full potential, such an organizational innovation should probably be comple- mented by other organizational innova- tions, such as the introduction of new human resource management practices or new software production techniques. Therefore, exploring the impact of col- laborations with the OSS community on the introduction of organizational inno- vations by entrepreneurial ventures

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 359

would be a worthwhile addition to this work.

Third, we measure innovation perfor- mance through survey questions that resemble questions from the Community Innovation Survey. Such a method of data collection is well established in the innovation literature, but we are well aware that measuring innovation is a challenging task in the software realm (for a discussion of this issue, see, e.g., Rossi-Lamastra 2009). Therefore, future studies on the relationship between col- laborations with the OSS community and firms’ innovation performance would benefit from the introduction of alterna- tive indicators of innovation perfor- mance (e.g., based on expert assessment and case study evidence).

Despite its limitations, our paper has relevant practical implications. First, our data show that OSS deserves the confi- dence placed in it by entrepreneurs and managers. High-tech markets are cur- rently globalized and highly competitive. Such hypercompetitive arenas not only magnify the traditional liabilities of rela- tively new ventures but also encourage them to engage deeply in the race for innovation. Our findings indicate that col- laborating with the OSS community may be a winning strategy for software entre- preneurial ventures to compete in the market. Our results may also be of interest to policymakers aiming to stimulate inno- vation by entrepreneurial ventures. OSS has been attracting the attention of poli- cymakers for years; however, to date, policy measures have largely been directed to support the adoption of OSS by public bodies (Rentocchini and Tartari 2010). The idea of incentivizing entrepre- neurial ventures’ collaborations with the OSS community is still lagging. By offer- ing evidence of the positive impact of collaborations with the OSS community on OSS entrepreneurial ventures’ innova- tion performance, the present study can provide a rationale for policy interven- tions in this area.

References Acs, Z. J., D. B. Audretsch, and M.

Feldman (1992). “Real Effects of Academic Research: Comment,” American Economic Review 82(1), 363–367.

Alexy, O. (2008). “Putting a Value on Openness: The Effect of Product Source Code Releases on the Market Value of Firms,” SSRN Working Paper, Available at http://ssrn.com/ abstract=1019527.

Alexy, O., and M. Leitner (2011). “A Fistful of Dollars: Financial Rewards, Payment Norms, and Motivation Crowding in Open Source Software Development,” European Manage- ment Review, forthcoming.

Antonelli, C. (1995). “The Diffusion of New Information Technologies and Productivity Growth,” Journal of Evolutionary Economics 5(1), 1–17.

Asquith, P., and D. W. Mullins (1986). “Equity Issues and Offering Dilution,” Journal of Financial Economics 15, 61–89.

Audretsch, D. B., and M. P. Feldman (1996). “R&D Spillovers and the Geo- graphy of Innovation and Produc- tion,” American Economic Review 86(3), 630–640.

Baron, J. N., and M. T. Hannan (2002). “Organizational Blueprints for Success in High-Tech Start-Ups,” California Management Review 44(3), 8–36.

Baron, R. A. (2010). “Job Design and Entrepreneurship: Why Closer Connections = Mutual Gains. Potential Contributions of Job Design to Entre- preneurship,” Journal of Organiza- tional Behavior 31, 370–378.

Becheikh, N., R. Landry, and N. Amara (2006). “Lessons from Innovation Empirical Studies in the Manufactur- ing Sector: A Systematic Review of the Literature from 1993–2003,” Technovation 26(5–6), 644–664.

Becker, S. W., and G. Gordon (1966). “An Entrepreneurial Theory of Formal

JOURNAL OF SMALL BUSINESS MANAGEMENT360

Organizations Part I: Patterns of Formal Organizations,” Administra- tive Science Quarterly 11(3), 315– 344.

Beckman, M. C., and M. D. Burton (2011). “Bringing Organizational Demography Back in: Time, Change and Structure in Top Management Team Research,” in Handbook of Top Management Team Research. Ed. M. Carpenter. Northampton, MA: Edward Elgar, 49–70.

Berger, A. N., and G. F. Udell (1998). “The Economics of Small Business Finance: The Roles of Private Equity and Debt Markets in the Financial Growth Cycle,” Journal of Banking and Finance 22, 613– 673.

Bessen, J., and R. Hunt (2007). “An Empirical Look at Software Patents,” Journal of Economics and Manage- ment Strategy 16, 157–189.

Bonaccorsi, A., S. Giannangeli, and C. Rossi (2006). “Entry Strategies Under Competing Standards: Hybrid Busi- ness Models in the Open Source Soft- ware Industry,” Management Science 52(7), 1085–1098.

Bonaccorsi, A., and C. Rossi (2003). “Why Open Source Can Succeed,” Research Policy 32(7), 1243–1258.

——— (2004). “Contributing to the Common Pool Resources in Open Source Software. A Comparison between Individuals and Firms,” SSRN Working Paper, Available at http:// ssrn.com/abstract=430920 or doi:10. 2139/ssrn.430920.

——— (2006). “Comparing Motivations of Individual Programmers and Firms to Take Part in the Open Source Movement. From Community to Busi- ness,” Knowledge, Technology and Policy 18(4), 40–64.

Bonaccorsi, A., C. Rossi, and A. Scateni (2005). “ELISS Report,” European Libre Software Survey, Mimeo.

Bryant, P. C., and D. G. Allen (2009). “Emerging Organizations’ Characteris-

tics as Predictors of Human Capital Employment Mode: A Theoretical Perspective,” Human Resource Management Review 19(4), 347– 355.

Cameron, A., and P. Trivedi (2005). Microeconometrics: Methods and Applications. New York: Cambridge University Press.

Capra, E., C. Francalanci, F. Merlo, and C. Rossi-Lamastra (2011). “‘Firms’ Involvement in Open Source Projects: A Trade-off between Software Quality and Success,” Journal of Systems and Software 84, 144–161.

Carpenter, R. E., and B. C. Petersen (2002a). “Is the Growth of Small Firms Constrained by Internal Finance?,” Review of Economics and Statistics 84, 298–309.

——— (2002b). “Capital Market Imper- fections, High-Tech Investment, and New Equity Financing,” Economic Journal 112, F54–F72.

Cassiman, B., and R. Veugelers (2006). “In Search of Complementarity in Innovation Strategy: Internal R&D and External Knowledge Acquisition,” Management Science 52, 68–82.

Colombo, M. G., L. Grilli, and E. Piva (2006). “In Search of Complementary Assets: The Determinants of Alliance Formation of High-Tech Start-Ups,” Research Policy 35, 1166–1199.

Colombo, M. G., and E. Piva (2008). “Strengths and Weaknesses of Aca- demic Start-Ups: A Conceptual Model,” IEEE Transactions on Engineering Management 55(1), 193– 206.

Colombo, M. G., E. Piva, and C. Rossi- Lamastra (2011). “Organizing for Col- laborating with Communities of Users and Developers: The Case of Firms Doing Business with the Open Source Software Community,” Working Paper, Politecnico di Milano.

Colombo, M. G., and C. Rossi-Lamastra (2011). “The Organizational Design of High-Tech Start-Ups: State of the Art

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 361

and Directions for Future Research,” Working Paper, Politecnico di Milano.

Connelly, B. L., S. T. Certo, R. D. Ireland, and C. R. Reutzel (2011). “Signaling Theory: A Review and an Assessment,” Journal of Management 37(1), 39–67.

Cooper, A. C., and A. V. Bruno (1977). “Success among High-Technology Firms,” Business Horizons 20, 16– 22.

Dahlander, L. (2007). “Penguin in a New Suit: A Tale of How De Novo Entrants Emerged to Harness Free and Open Source Communities,” Industrial and Corporate Change 16(5), 913–943.

Dahlander, L., and M. Magnusson (2005). “Relationships between Open Source Software Companies and Communi- ties: Observations from Nordic Firms,” Research Policy 34, 481–493.

Dahlander, L., and M. Magnusson (2008). “How Do Firms Make Use of Open Source Communities?,” Long Range Planning 41(6), 629–649.

Dahlander, L., and M. Wallin (2006). “A Man on the Inside: Unlocking Com- munities as Complementary Assets,” Research Policy 35(8), 1243–1259.

Eilhard, J. (2008). “Firms on Source- Forge,” MPRA Paper No. 7809.

Feeser, H. R., and G. E. Willard (1990). “Founding Strategy and Performance: A Comparison of High and Low Growth High-Tech Firms,” Strategic Management Journal 11, 87–98.

Feller, J., and B. Fitzgerald (2002). Understanding Open Source Software Development. Boston, MA: Addison Wesley.

Fitzgerald, B. (2006). “The Transforma- tion of Open Source Software,” MIS Quarterly 30(3), 587–598.

Folloni, G., and G. Vittadini (2010). “Human Capital Measurement: A Survey,” Journal of Economic Surveys 24(2), 248–279.

Gruber, M., and J. Henkel (2006). “New Ventures Based on Open Innovation—An Empirical Analysis of Start-up Firms in Embedded Linux,”

International Journal of Technology Management 4, 356–372.

Hargrave, T. J., and A. Van de Ven (2006). “A Collective Action Model of Institutional Innovation,” Academy of Management Review 31(4), 864– 888.

Hecker, F. (1999). “Setting Up the Shop: The Business of Open-Source Soft- ware,” IEEE Software 16(1), 45–51.

Heckman, J. (1978). “Dummy Endog- enous Variables in a Simultaneous Equations System,” Econometrica 46, 931–960.

Heckman, J., L. Lochner, and P. Todd (2003). “Fifty Years of Mincer Earn- ings Regressions,” NBER Working Paper (9732), NBER.

Henkel, J. (2009). “Champions of Revealing—The Role of Open Source Developers in Commercial Firms,” Industrial and Corporate Change 18(3), 435–471.

Hoskisson, R. E., J. Covin, H. W. Volberda, and R. A. Johnson (2011). “Revitalizing Entrepreneurship: The Search for New Research Opportuni- ties,” Journal of Management 48(6), 1141–1168.

Johnson, J. P. (2002). “Open Source Soft- ware: Private Provision of a Public Good,” Journal of Economics and Management Strategy 11(4), 637– 662.

Katsamakas, E. G., and N. C. Georgant- zas (2010). “Open Source Disruptive- Innovation Strategy,” Human Systems Management 29, 217–229.

Knockaert, M., D. Ucbasaran, M. Wright, and B. Clarysse (2011). “The Relation- ship between Knowledge Transfer, Top Management Team Composition, and Performance: The Case of Science-Based Entrepreneurial Firms,” Entrepreneurship Theory and Practice 35(4), 777–803.

Lakhani, K. R., and E. von Hippel (2003). “How Open Source Works: Free User- to-User Assistance,” Research Policy 32, 923–943.

JOURNAL OF SMALL BUSINESS MANAGEMENT362

Laursen, K., and A. Salter (2006). “Open for Innovation: The Role of Openness in Explaining Innovation Performance among UK Manufacturing Firms,” Strategic Management Journal 27(2), 131–150.

Lee, C., K. Lee, and J. M. Pennings (2001). “Internal Capabilities, External Networks, and Performance: A Study on Technology- Based Ventures,” Strategic Management Journal 22, 615–640.

Lepak, D. P., and S. A. Snell (1999). “Examining the Human Resource Architecture: The Relationships among Human Capital, Employment, and Human Resource Configurations,” Journal of Management 28(4), 517– 543.

Lerner, J., P. P. Pathak, and J. Tirole (2006). “The Dynamics of Open- Source Contributors,” American Eco- nomic Review 96(2), 114–118.

Lerner, J., and J. Tirole (2002). “Some Simple Economics of the Open Source,” The Journal of Industrial Economics 2(L), 197–234.

——— (2005). “The Economics of Tech- nology Sharing: Open Source and Beyond,” The Journal of Economic Perspectives 19(2), 99–120.

Mairesse, J., and P. Mohnen (2010). “Using Innovation Surveys for Econo- metric Analysis,” in Handbook of the Economics of Innovation. Eds. B. H. Hall and N. Rosenberg. Amsterdam: North-Holland, 1129–1155.

Mann, R., and T. W. Sager (2007). “Patents, Venture Capital, and Soft- ware Start-Ups,” Research Policy 36, 193–208.

McGowan, D. (2001). “Legal Implications of Open Source Software,” University of Illinois Law Review 1, 241–304.

McMullen, J. S., and D. A. Shepherd (2006). “Entrepreneurial Action and the Role of Uncertainty in the Theory of the Entrepreneur,” Academy of Management Review 31(1), 132– 152.

Miranda, A., and S. Rabe-Hesketh (2006). “Maximum Likelihood Estimation of Endogenous Switching and Sample Selection Models for Binary, Ordinal, and Count Variables,” Stata Journal 6(3), 285–308.

O’Mahony, S. (2003). “Guarding the Commons: How Community Managed Software Projects Protect Their Work,” Research Policy 32(7), 1179–1198.

O’Mahony, S., and B. A. Bechky (2008). “Boundary Organizations: Enabling Collaboration among Unexpected Allies,” Administrative Science Quarterly 53(3), 422–459.

Piva, E., L. Grilli, and C. Rossi-Lamastra (2011). “The Creation of NTBFs at the Local Level: The Role of Local Com- petences and Communication Infra- structures,” Industry and Innovation 18(6), 563–580.

Piva, E., F. Rentocchini, and C. Rossi- Lamastra (2011). “Absorbing Knowl- edge from Unconventional Sources. How Collaborations with the Open Source Community Shape the Innova- tion Performance of Entrepreneurial Ventures,” paper presented at the DRUID 2011 conference, Copenhagen Business School.

Rentocchini, F., and D. Tartari (2010). “An Analysis of the Adoption of Open Source Software by Local Public Administrations: Evidence from the Emilia-Romagna Region of Italy,” International Journal of Open Source Software and Processes 2(3), 1–29.

Romijn, H., and M. Albaladejo (2002). “Determinants of Innovation Capabi- lity in Small Electronics and Software Firms in Southeast England,” Research Policy 31(7), 1053–1067.

Rosen, L. (2001). “Which Open Source License Should I Use for My Soft- ware?” Available at: http://www. rosenlaw.com/html/GL5.pdf.

Rossi, C., and A. Bonaccorsi (2005). “Intrinsic Motivations and Profit- Oriented Firms in Open Source Soft- ware. Do Firms Practise What They

PIVA, RENTOCCHINI, AND ROSSI-LAMASTRA 363

Preach?,” in The Economics of Open Source Software Development Analyz- ing Motivation, Organization, Inno- vation and Competition in the Open Source Software Revolution. Eds. J. Bitzer and P. J. H. Schroeder. Amster- dam: Elsevier, 83–110.

Rossi-Lamastra, C. (2009). “Software Innovativeness. A Comparison between Proprietary and Free/Open Source Solutions Offered by Italian SMEs,” R and D Management 39(2), 153–169.

Stam, W. (2009). “When Does Commu- nity Participation Enhance the Perfor- mance of Open Source Software Companies?,” Research Policy 38(8), 1288–1299.

Stevenson, H. H., and D. E. Gumpert (1985). “The Heart of Entrepreneur- ship,” Harvard Business Review 63(2), 85–94.

Stuart, T. E., and O. Sorenson (2007). “Strategic Networks and Entre- preneurial Ventures,” Strategic Entrepreneurship Journal 1, 211–227.

Teece, D. J. (1986). “Profiting from Tech- nological Innovation: Implications for Integration, Collaboration, Licensing, and Public Policy,” Research Policy 15, 285–305.

Unger, J. M., A. Rauch, M. Frese, and N. Rosenbusch (2011). “Human Capital

and Entrepreneurial Success: A Meta- Analytical Review,” Journal of Busi- ness Venturing 26(3), 341–358.

Van de Vrande, V., J. P. J. de Jong, W. Vanhaverbeke, and M. de Rochemont (2009). “Open Innovation in SMEs: Trends, Motives and Management Challenges,” Technovation 29, 423– 437.

Van de Vrande, V., W. Vanhaverbeke, and O. Gassmann (2010). “Broaden- ing the Scope of Open Innovation: Past Research, Current State and Future Directions,” International Journal of Technology Management 52(3–4), 221–235.

von Krogh, G., C. Rossi-Lamastra, and S. Haefliger (2012). “Phenomenon- Based Research in Management and Organization Science: The Case of Open Source Software,” Working Paper, Long Range Planning, forthcoming.

West, J., and S. Gallagher (2006). “Chal- lenges of Open Innovation: The Paradox of Firm Investment in Open-Source Software,” R and D Management 36(3), 319–331.

West, J., and S. O’Mahony (2008). “The Role of Participation Architecture in Growing Sponsored Open Source Communities,” Industry and Innova- tion 15(2), 145–168.

JOURNAL OF SMALL BUSINESS MANAGEMENT364

Copyright of Journal of Small Business Management is the property of Wiley-Blackwell and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.