Intrapreneurship or Entrepreneurship?

Intrapreneurship or Entrepreneurship? I explore the factors that determine whether new business opportunities are exploited by starting a new venture for an employer (‘nascent intrapreneurship’) or independently (‘nascent entrepreneurship’). Analysis of a nationally representative sample of American adults gathered in 2005-06 uncovers systematic differences between the drivers of nascent entrepreneurship and nascent intrapreneurship. Nascent entrepreneurs tend to leverage their general human capital and social ties to organize ventures which sell directly to customers, whereas intrapreneurs disproportionately commercialize unique new opportunities which sell to other businesses. Implications of the findings are discussed. JEL Classification: L26, M13 Keywords: nascent entrepreneurship, intrapreneurship, sample selection Corresponding author: Simon C. Parker Richard Ivey School of Business University of Western Ontario 1151 Richmond St London, Ontario N6A 3K7 Canada E-mail: [email protected] Intrapreneurship or Entrepreneurship? 2 1. Introduction Intrapreneurship — also known as corporate entrepreneurship and corporate venturing (Burgelman 1983; Burgelman 1984; MacMillan et. al. 1986) — is the practice of developing a new venture within an existing organization, to exploit a new opportunity and create economic value (Pinchot 1985). Entrepreneurship by contrast is the act of developing a new venture outside an existing organization. There is evidence that intrapreneurship helps managers to renew and revitalize their businesses, to innovate, and to enhance their overall business performance (Kuratko et. al. 1990; Antoncic and Hisrich 1990). It is therefore of interest to investigate which factors are associated with e¤orts to start new ventures for an employer by ‘nascent intrapreneurs’, and which are associated instead with independent start-ups by ‘nascent entrepreneurs’— where the word ‘nascent’refers to emergent (i.e., not yet completed) start-up activity. The present article explores this question. Previous research has identi…ed several reasons why new opportunities might be exploited via entrepreneurship rather than intrapreneurship. These include agency costs which a¤ect contracting between employees and employers; transferable human capital and limited asset complementarity within existing …rms; and organizational limitations of incumbents such as bureaucracy and rigid routines (Henderson 1993; Anton and Yao 1995; Klepper 2001; Helfat and Lieberman 2002). However, to date there is only limited empirical evidence about the factors promoting entrepreneurship rather than intrapreneurship. Instead, the evidence base is mainly concerned with: linking the market entry strategies and performance outcomes of new spino¤s to the performance of parent …rms (Christensen 1993; Helfat and Lieberman 2002; Gompers et. al. 2005; Klepper and Thompson 2006); organizational factors which in‡uence intrapreneurship (Pinchot 1985; Hornsby et. al. 2002); the e¤ects of intrapreneurship on corporate performance (Lumpkin and Dess 1996; Zahra and Gravis 2000); and characteristics of intrapreneurs (Fayolle 2004; Howell and Higgins 1990a). This article develops theoretical arguments and testable hypotheses about the factors promoting nascent intrapreneurship relative to nascent entrepreneurship. These arguments principally draw on contributions from human capital theory and incentive (agency) theory. The article utilizes data from a relatively large and representative publicly–available survey of American adults, the PSED Intrapreneurship or Entrepreneurship? 3 II (www.psed.isr.umich.edu). This dataset makes a clean distinction between NE and NI startups, and contains a rich array of explanatory variables about individuals and their products; the processes they follow; and the organizational environments they operate in, including information about their current or former employers (Aldrich 1999; Menzel et. al. 2007). The PSED II also contains information about people who choose neither intrapreneurship nor entrepreneurship. This makes it possible to deal with the possibility that nascent intrapreneurs and entrepreneurs self-select non-randomly into these activities. Dealing with sample selection turns out to be crucial for deriving accurate empirical results. The …ndings reveal a central role for general human capital, which promotes start-up organizing e¤orts in general, and nascent entrepreneurship (rather than nascent intrapreneurship) in particular. While previous research has already identi…ed general human capital as an important determinant of start-up e¤orts in general (e.g. Davidsson and Honig 2003), its in‡uence on NE relative to NI is less well known. Furthermore, it is found that independent starts are more likely to exploit ideas which involve talking to customers rather than researching established competitors, and which leverage social ties as well as general human capital. In contrast, incumbent …rms tend to pursue intrapreneurship by developing unique business-to-business products which seem to be harder for employees to transfer into independent start-up e¤orts. We believe that these results are not only interesting in their own right but also could prove informative for theory and practice. For example, because the (observable) characteristics of individuals, their current or more recent workplaces, and their proposed new products all turn out to a¤ect the decision to do NI or NE, future theories might need to be enriched to draw on a broader range of considerations than has been the case hitherto. In particular, our …ndings suggest the need for future theory to extend the scope of inquiry from that of the individual employee to take greater account of the objectives, capabilities and limitations of the organizations they work for. Strategic interactions between agents in this context are likely to be both interesting and wide-ranging. Future theory development might also fruitfully delve deeper into the self-selection mechanisms (based on unobservable characteristics) which our empirical investigation uncovers. Intrapreneurship or Entrepreneurship? 4 The next section develops theoretical arguments and testable hypotheses about the factors promoting NI relative to NE. The section that follows it explains the dataset, including key explanatory variables and control variables. The …fth section outlines the empirical methods used to test the hypotheses. The sixth section presents the results. The …nal section concludes. 2. Theory A large body of theoretical research emphasizes the importance of human capital for understanding the determinants of start-up organizing e¤orts. It is convenient at the outset to follow Becker (1964) by distinguishing between ‘general’and ‘speci…c’human capital. General human capital comprises skills, knowledge, experience and capabilities (such as those embodied in formal education) which are useful in a multitude of productive uses, including both existing organizations and new venture creation. Speci…c human capital in contrast refers to skills, experience, knowledge and capabilities, such as those imparted by …rm-speci…c training programs, which are primarily useful to the organization which provides them. The distinction between general and speci…c human capital helps to elucidate implications of an agency problem studied by Anton and Yao (1995), in which an employee identi…es a new venture opportunity which can be exploited either inside or outside the …rm. The employee can keep the opportunity secret, and quit the …rm in order to exploit it in a new independent …rm (NE); or she can disclose it to the …rm in the hope of sharing in the pro…ts resulting from joint development (NI). The greater an employee’s general human capital, the greater is their capability to exploit the opportunity outside the …rm in independent NE (Zucker et. al. 1998). General human capital is associated with analytical ability, knowledge about business opportunities and conditions, and computational and communication skills which are readily transferable between organizations. In contrast, an employee’s general human capital may be less relevant for developing the opportunity via NI if the …rm can substitute other workers’general human capital to develop the idea once it has been revealed to them. Hence one might expect general human capital to be more associated with NE than with NI activity. Intrapreneurship or Entrepreneurship? 5 The value of new venture opportunities discovered within …rms is frequently enhanced by complementarities with other assets speci…c to the …rm (Teece 1986; Subramanian 2005). For example, an employee’s speci…c human capital can a¤ect the development trajectory of a new idea, culminating in an innovation which is complementary to the internal organization of the …rm and hence more valuable if exploited within it. In such cases, intrapreneurship would appear to be a logical option. But although speci…c human capital might be di¢ cult to transfer outside incumbent …rms owing to formal barriers (e.g. non-compete agreements) and informal barriers (e.g. embedded organizational skills), it is less obvious which aspects of speci…c human capital are truly speci…c to the …rms they are acquired in. Indeed, several theories of independent spino¤s argue that ‘speci…c’human capital is actually more transferable outside the …rm than it …rst appears (Helfat and Lieberman 2002; Møen 2005). Consider, for example, knowledge about a …rm’s internal organizational routines, which is an important kind of speci…c human capital (Nelson and Winter 1982; Klepper 2001). Although this knowledge might be redundant in other incumbent …rms, employees leaving a …rm to start a new independent venture might be able to replicate and transfer these routines to the new venture to better exploit their opportunities (Klepper 2001; Freeman and Engel 2007). In which case, what appears to be speci…c human capital might actually promote NE as well as NI activity, having more ambiguous e¤ects on choices of commercialization mode than general human capital does. These ideas are summarized in the following hypothesis: Hypothesis 1. General human capital makes NE more attractive relative to NI, and has stronger e¤ects on the mode of commercialization than speci…c human capital. Other …rm-speci…c assets include established marketing, sales, design and logistics departments, which are set up to service a …rm’s existing operations but which can also work with a new internal product line. The existence of these assets; a reputation for providing a stable and trusted source of supply; and greater legitimacy relative to new independent start-ups, give incumbent …rms an advantage over new independent ventures in commercializing new opportunities, whether of Intrapreneurship or Entrepreneurship? 6 the business-to-business (B2B) or business-to customer (B2C) sort. However, it seems likely that incumbent …rms have a comparative advantage over new ventures in B2B rather than B2C sales (Claycomb et. al. 2005). The reasoning is as follows. Few independent start-ups have access to the resources needed to supply other businesses (especially legitimacy and reputation), and instead are likely to have a comparative advantage in forging B2C relationships whose e¤ectiveness depends on ‡exible adjustment to fast-changing consumer demand. In these circumstances, incumbents’ routines can be a positive disadvantage, since redeploying resources from their current uses can cause ine¢ ciencies as managers of incumbent …rms strive to retain those resources to support the projects for which they are responsible. This can allow independent entrepreneurship to exploit pro…table (though possibly short-lived) niches (Audretsch and Thurik 2001; Freeman and Engel 2007). Hypothesis 2. Business-to-business opportunities are associated more with NI than NE commercialization: business-to-customer opportunities are associated more with NE than NI. Anton and Yao (1995) argue that greater personal wealth enables an employee to contract more e¤ectively with their employer. Anton and Yao (1995) consider a contract under which, in return for posting a bond to the employer, an employee who chooses joint commercialization is guaranteed a stake of the pro…ts. This contract removes the risk of expropriation by the employer and makes the employee willing to reveal their discovery — so promoting NI commercialization. In contrast, employees lacking wealth cannot post a bond. Being exposed to the threat of expropriation, they are likelier to take their opportunity outside the …rm, via the NE route. Hence wealth may be necessary for NI exploitation. A di¤erent argument proposes a positive relationship between available income and the probability of the NI mode. Managers wishing to discourage employees from quitting to found potentially competing start-ups might seek to ‘buy them o¤’with higher salaries, which helps retain both the employee and the innovation within the …rm (Møen 2005; Subramanian 2005; Hvide and Kristiansen 2007; Hvide 2009). More generally, it has been argued that even when employers lack Intrapreneurship or Entrepreneurship? 7 complementary assets and property rights, they can still design compensation contracts which successfully retain employees with valuable new ideas within the …rm (Pakes and Nitzan 1983; Anton and Yao 1994; Møen 2005). For these reasons, I propose the next hypothesis: Hypothesis 3. Greater personal wealth and income make NI more attractive relative to NE. The types of products commercialized by new start-ups are also likely to di¤er between NE and NI modes. Large …rms can spread the risks of innovation, giving them greater incentives to engage in the risky activity of radical innovation. They also perform more R & D (Cohen and Klepper 1992), which is often needed to pioneer new innovations. Both of these arguments suggest that NI is associated with unique products. Incumbents also have greater incentives to innovate in order to retain monopoly pro…ts than new independent entrants, since the latter can at best obtain duopoly pro…ts by competing with incumbents (Gilbert and Newbery, 1982). Furthermore, an incumbent is more likely to support a corporate venture innovating a unique new product which does not compete with (i.e. cannibalize the pro…ts from) the incumbent’s existing product line. And a new product presumably has to be su¢ ciently di¤erent from a …rm’s existing product line in order for it to be commercialized as a new business unit (NI) rather than via simply internal expansion. Another set of incentives is pertinent for managers who try to motivate their employees to work on a main task. Managers can often do best by committing to a policy of passing up mundane innovations discovered by employees to keep them focused on their main task; it only pays managers to break this policy rule and back NI when unique, valuable projects emerge (Hellmann 2007). On the other hand, unique new products can create their own problems for corporations. They can disrupt valuable organizational routines which work smoothly and e¤ectively for the …rm’s existing product lines. And they can threaten to render obsolete the expertise underlying the careers of managers responsible for deciding the innovation strategy of the …rm. As a result, managers may prefer to back relatively predictable incremental innovations rather than ambitious radical ones (Bhide 2000; Freeman and Engel 2007). Furthermore, uncertainty associated with unique new products can promote divergent opinions between employees and employers about the value Intrapreneurship or Entrepreneurship? 8 of a discovery. That makes it less likely that both parties will agree on initiating a corporate venture (Audretsch 2001; Klepper and Sleeper 2005). Hence managers wishing to encourage NI may need to obtain objective information to convince sceptics within their organizations about a new corporate venturing opportunity, especially if it is based around a unique new product. In particular, managers may need to acquire information about competitors who might also be working on similar developments. With this information in place, incumbent …rms should be more willing to adjust routines and resolve internal con‡icts to facilitate intrapreneurship. Hence we might expect research about competitors to promote NI commercialization directly, as well as to moderate positively the relationship between the uniqueness of a venture’s product and the likelihood of NI commercialization. These arguments are summarized in the next hypothesis: Hypothesis 4. Products or services with unique attributes are more likely to be developed via NI than NE. NI exploitation of these opportunities entails research about competitors, which both promotes NI commercialization directly and moderates positively the relationship between the uniqueness of a venture’s product and the likelihood of NI commercialization. Entrepreneurship researchers have established that people who are engaged in start-up activities, whether NE or NI, are not representative of the population but are instead a self-selected group, amounting to no more than 6% of the American adult population (Gartner et. al. 2004; Reynolds et. al. 2004). If these self-selected individuals have unobserved attributes which predispose them to a particular mode of entry, then it is important to take this into account when analyzing empirically NI–NE outcomes. For example, consider two hypothetical individuals, A and B, where A has unobserved attributes which predispose them to entrepreneurial-type work over corporate wage-and-salary work, unlike B who prefers the corporate environment. It seems plausible given these preferences that B is more likely (all else equal) to engage in NI if they try a start-up at all. In contrast, independence-seeking A is more likely (all else equal) to engage in NE if they try a