مسیر ناهموار مشارکت های فن آوری: درک علل و عواقب ناشی از مشارکت سوء عملکرد

Research on technological partnerships has traditionally sought explanation of their high failure rates in partner characteristics and relationship features. This study introduces the notion of a ‘bumpy road’ in technology partnerships which refers to undesired outcomes such as ‘partnership mal-functioning’ and ‘instability’ to the degree to which innovation activities are hampered. We explain how firm-level strategies can reduce the probability of a ‘bumpy road’ in partnerships. We also assess the impact of this ‘bumpy road’ on innovative performance. We find that firms that excel in diversification of external activities (in terms of different types of partners) perform best. Moreover, a persistent product oriented innovation strategy geared at developing new products, new markets, or higher product quality will yield more stable partnership outcomes. Our results confirm that engagement in partnerships is beneficial for innovative performance. However, firms that experienced a ‘bumpy road’ in their technological partnerships have to pay a price in terms of a negative effect on their innovative performance.

Failure is a frequent outcome of inter-firm partnering (Park and Ungson, 2001). High failure rates of 30–50% in partnerships are not an uncommon finding in the literature (e.g., Bleeke and Ernst, 1991, Harrigan, 1988a, Harrigan, 1988b and Killing, 1988). Especially technology partnerships, where firms’ exchange technology and jointly perform R&D in the context of complex intellectual property regimes, are inherently difficult to manage and hence this type of partnership is subject to the highest rates of failure (see, for example, Sadowski et al., 2005). Partnerships can be unsuccessful for various reasons. Prior research sought explanations for this phenomenon from the perspective of several theoretical approaches (see Das and Teng, 2000a for a literature overview). Transaction cost economics stresses the pursuit of self-interest at the expense of the partner as well as the high costs of deterring such opportunistic behavior as a major cause of partnership instability (Williamson, 1985 and Gulati, 1995). Game-theoretic approaches emphasize the role of uncertainty in predicting the intentions of partners and future payoffs (Parkhe, 1993a). The resource-based view suggests that inequality in resources that firms bring into a partnership gives rise to an eventual power imbalance between partners that can lead to a premature termination of their partnership (Bucklin and Sengupta, 1993, Das and Teng, 2000b and Osborn and Baughn, 1990). From the perspective of a strategic behavior approach, the literature points at the role of inter-firm rivalry and competition that increases the likelihood of partnership instability (Porter, 1985, Porter, 1990 and Kogut, 1989). The literature also indicates that partnerships are motivated by the need to share both risks and costs of R&D, to gain access to new technology and new markets, and to create synergetic effects (Hagedoorn, 1993). However, successful partnerships between competitors are expected to be rare because often achievement of these goals proves unrealistic, leading to a premature termination of a partnership (Porter, 1990). Partnerships with substantial overlap in core businesses, geographic markets, and functional skills are reported to have success rates of about 30% as competitors are inclined to maximize their own individual objectives rather than their collaborative interests. Moreover, managerial complexity of these joint activities induces a higher probability of partnership failure (Park and Ungson, 2001). In a partnership, two or more independent firms need to be coordinated and also the activities of the partnership have to be aligned with the parent firms’ interests. This coordination may be cumbersome due to, for instance, cultural and organizational differences between partners turning partnership demise into a likely outcome (Parkhe, 1993b). So far, the existing literature has seldom adhered to a uniform terminology when describing partnership outcomes. When assessing the positive performance of partnerships objective performance measures (e.g., sales growth, profitability, and return on assets) can be used or, alternatively, survey data can be employed where managers are asked directly to assess the performance of a partnership. For the assessment of less successful partnership outcomes, terms like ‘failure’, ‘premature termination’, ‘dissolution’ and ‘instability’ have been used interchangeably to indicate unfavorable results of a partnership (see Das and Teng, 2000a and Park and Ungson, 2001 for overviews). In our study we take a combination of these two approaches as we consider both the functioning and the results of technology partnerships. We employ a measure of firms experiencing a ‘bumpy road’ in their technology partnerships related to the notion of partnership instability as discussed by Das and Teng (2000a). We focus on firms that experienced a ‘bumpy road’ in their technology partnerships when unplanned outcomes such as stoppage, delay or abandonment of their joint innovation project(s) occurred. A somewhat similar approach has been recently adopted in Lhuillery and Pfister (2009). Our analysis deviates from their approach in a number of ways. First, we intentionally adopt a broader definition than simply premature partnership dissolution, by also analyzing those events when firms had difficulty-laden or even unproductive partnerships. We do so to explore a broader issue of what factors and firm strategies can reduce the probability of things going wrong in partnerships. Cooperation that results in a failure is only one instance of this broader mal-functioning of partnerships. Second, in the context of this broader perception of the mal-functioning of partnerships as their ‘bumpy road’, we investigate whether firms that employ a persistent innovation strategy geared at either reducing cost or developing new products face a lower probability of such a ‘bumpy road’ in their technology partnerships. Third, we consider whether there is an inverse relationship between the diversity and breadth in the partnership portfolio of firms and the probability of their partnership mal-functioning. Finally, in our analysis we look at the effect that the ‘bumpy road’ in the technological partnerships of firms might have on their future innovative performance. In the following, we study persistence in the overall firm's business strategy with respect to firm's innovative activities (product oriented vs. cost oriented) and in terms of firm's alliance strategy. Our approach follows the definition of persistence as ‘state dependence’ (e.g., Heckman, 1981), which in our context means that firms are studied from the perspective of both their past innovation activities/partnerships and their current continuation of these activities. A similar approach has been used to analyze persistence in profits (Mueller, 1977 and McGahan and Porter, 2003), innovation (Roberts, 1999 and Raymond et al., 2010) or other measures of firm performance such as Tobin's q (Villalonga, 2004). In these studies, persistence is also understood in terms of a relation between the ‘current state’ of a firm (in terms of its strategy or performance) and the ‘past states’ for the same dimension of its activities. To test our hypotheses we employ three consecutive streams of data for the period 1994–2000 collected through the Community Innovation Survey (CIS). The CIS survey is organized by Eurostat and is aimed at collecting information on firms’ innovating activities. The survey covers both large R&D performing firms as well as smaller innovating firms with a limited number of technology partnerships. The paper proceeds as follows. The next section presents the theoretical framework and derives hypotheses. We then describe the data, variables and methods. This is followed by a presentation of the results, the final section discusses these results and it presents the conclusions.

Technology partnerships are inherently difficult to manage and as a result partnership instability rates are notoriously high. Our empirical results indicate that it is costly to encounter serious problems in technology partnerships. We find that for partnering firms going through a ‘bumpy road’, i.e. when badly performing technology partnerships induce unplanned outcomes as stoppage, delay or abandonment of innovation projects, the innovative performance is reduced substantially. In particular, in terms of generating new sales through new products experiencing a mal-functioning of technology partnerships is very costly. As a consequence, we think that firms have an incentive to counter such problems effectively. Our results indicate that firms may pursue several strategies to reduce the likelihood of a ‘bumpy road’ in their partnerships. First, we find that firms that persistently rely on a product focus innovation strategy through developing new products, new markets and improving product quality face better chances in dealing with a ‘bumpy road.’ A strategy based on the exploration of these new opportunities, where firms persistently engage in joint innovation projects, rather than jointly exploiting cost efficiencies, lowers the probability of inter-organizational mal-functioning. As discussed in the above, persistent product and market innovators, are more successful in avoiding partnership problems because they face less uncertainty regarding both the goals of an alliance and the appropriate partnership candidate. In addition, firms with an innovative reputation, are desirable partners for technology partnerships (Stuart et al., 1999) and this reputation limits the potential frictions in technology partnerships. This result for the effect of a persistent innovation strategy, as a signal of consistency to partners, also indicates that it is important to avoid changing the firm's priorities in order to prevent partnership malfunctioning. Second, we find that generating a diverse portfolio of partnerships may be helpful in obtaining the required experience in dealing with partners. Our results corroborate existing field research that has pointed out that firm's first-hand experience with partnerships and firm's ability to learn from its experience helps it avoid pitfalls in future partnerships (Lhuillery and Pfister, 2009 and Kale et al., 2002). In particular, our results indicate that firms with links to a diverse group of other firms are confronted with a much lower incidence of problematic partnership outcomes. This relevance of the diversity of partnerships can be linked to attempts of firms to avoid the pitfalls of local search constraints (Rosenkopf and Nerkar, 2001). A diversity of partnerships gives access to a range of tacit skills, many of which are related to the technological capabilities and knowledge sources of these partners (Kogut and Zander, 1992 and Lim, 2004). Firms that operate in a context of diverse contacts are more likely to benefit from an exposure to different ideas, varied experiences and a range of events through which they develop a more elaborate knowledge base on how to successfully manage their technology partnerships. Our findings suggest it is necessary to distinguish between the benefits of a current diverse portfolio and a diverse portfolio established in the past. We find that the current management of a diverse portfolio does not affect the probability of experiencing partnership malfunctioning. The learning experience materializes when a diverse portfolio was established some time ago. If some time has elapsed since the experience of a diverse partnership portfolio, capabilities can be generated that are instrumental for the effectiveness of future technology partnerships. However, these learning advantages have not yet materialized when the current portfolio of partnerships is very diverse. In particular, one may expect that if the current portfolio consists of a wide variety of different types of partnerships the management of this portfolio becomes more troublesome which may increase the probability of a ‘bumpy road.’ Third, the more firms are able to avoid ‘bumpy roads’ the higher their future innovative performance. As also established by Spender (2007) and Zahra and George (2002), firms with a higher diversity in their partnerships portfolio gradually develop skills that enable them to improve their ability to absorb valuable information from outside sources which we interpret as an ability to avoid mal-functioning in their partnership formation. The better organized the portfolio of partnerships, in the sense of avoiding troublesome partnerships, the better firms are at leveraging their newly acquired skills into new innovative activities that improve their innovative performance. Interestingly, these results for the technology partnering behavior of firms resonate the findings of a classical study on the difference between successful and unsuccessful innovators (Rothwell et al., 1974). Similar to our findings on the ‘bumpy road’ of technology partnerships and their role in the innovation performance of firms, Rothwell et al. (1974) found, among other things, that firms with a persistent innovation strategy and a more extensive use of outside resources were more successful innovators than firms that scored lower on these characteristics. In the current world, where innovation is much more embedded in inter-firm technology partnerships, our findings illustrate that a diversity of outside technological resources through a seasoned and diverse portfolio of partnerships, in the context of a product focused innovation strategy, is crucial if firms attempt to minimize the effects of malfunctioning technology partnerships. 5.1. Limitations and future research CIS follows a long tradition of Yale and SPRU innovation surveys by collecting a large amount of primary information on firms’ innovative strategies. The advantage of CIS compared to other surveys is that it is a large-scale survey that does the sampling and data collection in a rather rigorous way (in many countries it is administered by the national statistical bureaus) and that it covers in a representative way most of the ISIC 3 sectors and firms of different size classes. While offering researchers interesting insights into firm's innovation process, a large-scale database like CIS has a number of limitations. First, it puts a constraint on the researcher in terms of the measures that can be used. Some fine-grained information, for instance, about firms’ capabilities and the process that firms go through while organizing their partnerships, is not available. As pointed out earlier, one limitation of our framework is that we have no information on the relative importance of firms’ individual collaborations within their portfolio of partnerships. Information on the number of ties of each type is also not available. Second, although CIS is conducted on a bi-annual basis, some questions do not belong to the ‘core questionnaire’ and thus are rotated or deleted from one survey to another. In our study we worked with some questions that are rotated. While it did not hamper our ability to trace firms across a number of years, it did put limitations to our ability to construct a proper panel and control for firm-specific effects. In this paper we documented the benefits of persistent innovation strategies for alliance outcomes. Such persistency (as state dependence) can be due to various reasons such as resource needs, governance preferences, inertia and habit, path-dependency or the ‘remediableness criterion’ as discussed by Williamson (1996). A proper investigation of why persistence in a firm's innovation strategies arises in the first place is beyond the scope of this paper and is left for further research. Another interesting avenue for future research is a further investigation of the bi-directional link between firm's partnership capabilities and the performance of its technological partnerships. Initial exploratory research shows that some capabilities are more important than others and that firms can enhance partnership performance through development of particular partnership capabilities, which may act as a mediator between partnership experience and partnership performance (Zollo and Winter, 2002 and Heimeriks and Duysters, 2007). A somewhat related avenue is the (mediating) effect of organizational competencies and specific managerial practices on firm's partnership performance. In our framework we highlighted the benefits of pursuing certain firm-level innovation strategies to reduce the probability of partnership-malfunctioning, yet a better understanding of the interplay between organizational and strategic aspects may be needed.