Towards A Meta-framework of Collaborative Learning

Figure 27.2 presents our mapping of the field. We discuss its dimensions in the following section.

Figure 27.2 Mapping the Collaborative Field

Alliances: specifying the boundaries

A range of inter-firm relationships can be categorized as strategic alliances (Kale and Singh, 2009). Definitional ambiguities render assessing cooperation, particularly, inter-firm cooperation, difficult. Early on, O’Brien and Tullis (1989), Root (1988) and Shenkar and Zeira (1987) recognized that the term ‘strategic alliances’ already was becoming little more than a buzzword. These definitional ambiguities are a major source of difficulty when comparing the results of different studies (Terpstra and Simonin, 1993). In light of confusion created by this proliferation of terms, we favor the use of the more universal and inclusive term ‘inter-organizational collaboration,’ or simply ‘alliances.’ Although many definitional propositions exist, their practical value may be quite limited. Rather, researchers should focus on the key variables that help frame the boundaries of collaboration. In Figure 27.2, these key variables are: alliance form, mode, governance, scope, number of partners, and cycle. Alliance form (Terpstra and Simonin, 1993) refers to the structural organization of the alliance along an equity continuum from no equity involved to full equity participation in the case of an acquisition or a merger. Past research has focused essentially on three forms: contractual agreements (e.g. licensing, franchising; no equity is involved), equity participation (e.g. equity swaps and partial acquisition; here, no new legal entity is created), and JVs (formation of a new and separate legal entity).

At the extremes of the equity continuum, two other forms of collaboration deserve attention. First, informal arrangements correspond to the case when no equity, but also no contract, is involved (e.g. exchange of personnel, benchmarking). Often unnoticed in the literature, this form of cooperation appears to be widely used. For instance, Berardo (2009) looks at cooperative funding programs where partnering governmental and nongovernmental organizations relate to each other essentially through informal links. In the corporate context, Hakanson and Johanson (1988) report in a study of cooperation between firms involved in technical development that more than two-thirds of the arrangements were informal. Second, the case of full equity (as opposed to no and partial equity) represents another, often ignored, case of cooperation. Mergers and acquisitions thus represent an ultimate form of collaboration where partners fully fuse their structures, legal existence, processes, cultures, and knowledge platforms. Research on knowledge acquisition and transfer in this area remains very promising and complementary (Zou and Ghauri, 2008; Westphal and Shaw, 2005; Henrik Bresman, Birkinshaw, and Nobel, 1999).

Similar to mergers and acquisitions, the multinational organization’s network of subsidiaries and affiliates fits our expanded definition. Lately, this area has seen a tremendous surge of research interest aimed at understanding knowledge management and flows, the epitome of the learning organization (Foss and Pedersen, 2004; Minbaeva, 2005, 2007; Monteiro, Arvidsson, and Birkinshaw, 2008; Oddou, Osland, and Blakeney, 2009; Simonin and Ozsomer, 2009). Over two decades ago, Hamel, Doz, and Prahalad (1989) warned that ownership structure seems to capture managers’ attention when, in fact, the learning place between partners is more important. Today, due to the growing interest in the learning organization and knowledge management, this impetus may have shifted. When looking at the alliance form, more is in play beyond the desire for control. The different structural forms of cooperation just identified may vary in their conduciveness to learning and types of prominent learning issues.

As form refers to the structural component of an alliance, mode focuses on the function of the collaboration (Terpstra and Simonin, 1993). These modes fall into two general categories Type X and Type Y (Porter and Fuller, 1986). Type X corresponds to joint activities (partners perform activities at the same level of the value chain; e.g. joint research and development) while Type Y relates to complementary activities (partners perform activities at different levels of the value chain; e.g. one partner provides the manufacturing capability whereas the other provides marketing). From a learning point of view, Type X is more propitious to maintenance and single-loop learning due to the likely presence of a common knowledge base. Type Y alliances may be sources of greater knowledge gaps between partners resulting in possible shifts of expertise (accumulation of radically different knowledge, double-loop learning), but, conversely, it is less likely that learning occurs due to the specialization and partitioning of knowledge across partners.

Next, governance refers to governance mechanisms (i.e. concrete management and control activities). We distinguish between two situations based on two common types (e.g. Hoetker and Mellewigt, 2009): (1) formal governance mechanisms that rely on the use of explicit agreements, indicators, reporting, and contracts to specify parties’ roles, expected performance, and dispute resolution means; and (2) relational governance mechanisms that foster open communication, sharing of information, trust, and cooperation. Much research interest exists in this domain (Faems, Janssens, Bouwen, and Van Looy, 2006; Kok and Creemers, 2008; Hoetker and Mellewigt, 2009) that is paralleled by extensive work in the area of ‘knowledge governance.’ Influential in this area is the collective work emanating from the Copenhagen Business School (e.g. Foss, 2007; Foss and Pedersen, 2004; Pedersen and Mahnke, 2004) that has put some emphasis, in particular, on the question of micro-foundations and processes (Abell, Felin, and Foss, 2008; Felin and Foss, 2009) and has drawn attention to the role of HRM (Minbaeva, 2005; Minbaeva, Foss, and Snell, 2009; Simonin and Ozsomer, 2009). Ultimately, these domains of investigation intersect and specialized research on knowledge governance in strategic alliances starts to emerge (Knudsen and Nielsen, 2009).

Finally, geography and activities define the scope of an alliance. In terms of geographical scope, one should differentiate domestic from international alliances on the basis of the national and linguistic origin of the partners (Meschi, 1997) and the location of the alliance itself or its interface (home or overseas) particularly when the focus is on knowledge transfer. The role of ‘cultural’ is intrinsically linked to the issue of location. Much has been written about the effects of national and organizational culture in alliances (Barger, 2007; Meirovich, 2010). The process of knowledge transfer, itself, has been shown to be influenced not just by the nature of competitive regime (partners as competitors or not), the form of the alliance (equity versus non-equity), but also by the degree of organizational distance across partners (Simonin, 2004). Recently, this line of inquiry has also drawn a lot more research interest when applied to related cases of knowledge transfer in MNEs, between headquarters and subsidiaries (e.g. Welch and Welch, 2008; Lucas, 2006).

Turning to the scope of alliance activities, one can distinguish the type of activities contributed (e.g. marketing, product development), their relatedness to the partners’ expertise and core competencies (central or peripheral), and the valence of their overall novelty (e.g. breakthrough or routine). The number of partners can influence learning processes and utilization. The case of two partners differs from cases with three or more partners because, in general, the degree of complexity in interactions grows in a non-linear way. At times, in specific collaborative context, the reverse may be true as well. For instance, when studying funding initiatives with government actors, Berardo (2009) found that collaborative efforts are more likely to succeed when the leading organization can secure the assistance of a larger number of partners. While most alliance research deals with dyadic relationships, studying consortia-like arrangements, particularly when assessing learning outcomes, processes, and performance levels, would contribute to the field.

Finally, when specifying alliance boundaries, one must isolate the components of an alliance cycle. Different challenges and problems exist at different stages of an alliance life (e.g. infant versus mature stage). Cooperation evolves and co-evolves over time and requires re-examination at various stages (Iyer, 2002; Duso, Pennings, and Seldeslachts, 2010). Duration is also an important aspect. Over time familiarity can increase, trust may intensify, and the opportunities for accessing knowledge might increase while the system diversity might decrease (Meschi, 1997; Gulati, 1995). Lastly, repetition provides another facet of a collaborative cycle. It encapsulates the degree and frequency at which partners collaborated in the past: is the alliance a first encounter between specific partners or, rather, another episode of a long collaborative history between them.

Learning: specifying the unit of analysis

Our model of collaborative learning identifies four distinct units of analysis: individuals, teams, organizations, and networks. These different levels of analysis also correspond to different learning foci and outcomes that are interrelated. This approach is consistent with Tiemessen, Lane, Crossan, and Inkpen’s (1997) organizational learning framework that recognizes three levels of learning (individual, group, organization). The alliance literature needs investigation into how to reconcile individual and organizational learning. Likewise, more research pertaining to knowledge structures and processes, with networks taken as a unit analysis, is needed. Overall, paying attention to units of analysis equates to more precision in the coverage and understanding of collaborative learning.

Organization: specifying the key characteristics

Beyond drawing alliance boundaries and the units of analysis, specifying key organizational characteristics that help further classify alliances and identify boundary conditions for learning outcomes and processes is necessary. Our model focuses on three categories: sector, type, and resources. Sector is a reminder that alliances go beyond business-to-business collaborations. Of particular interest, alliances between nonprofit organizations, public agencies, and hybrids with business organizations deserve greater attention (Austin, 2000). Under type, we regroup variables that characterize an organization and are likely to impact learning (e.g. age, size, nationality, industry, structure, culture, etc.). Finally, resources capture aspects of an organization pertaining to knowledge (competencies, technology, intangible assets).

Collaboration: specifying the locus, modus, focus, and orientation

Finally, we turn to four other key collaborative dimensions: collaborative locus, modus, focus, and collaborative orientation. Collaborative locus draws a distinction between inter- and intraorganizational collaborations. The bulk of alliance research deals with inter-firm collaborations. Our contention is that they are only a subset of the overall collaborative phenomenon. In a multinational enterprise, for instance, foreign subsidiaries or research teams from different SBUs collaborate with one another on specific projects. Best practices and global campaigns need to be shared within a network of affiliates. In a sense, we argue that the study of organizational learning and knowledge transfer in the context of multinationals and FDI is not so different from that in the context of traditional alliances: both fit under our model of collaborative learning. In one case the knowledge seeker may be a subsidiary in Spain and the knowledge holder/provider another subsidiary in Germany or HQ; in the other case (traditional alliance) two unrelated companies play these roles.

Collaborative modus corresponds to the degree of physical proximity and interaction between partners ranging from virtual collaboration to close physical proximity. As examined earlier, virtual collaboration has become a more prevalent phenomenon and area of research interest. Collaborative focus captures the degree to which an alliance falls under an ‘exploitation’ versus ‘exploration’ categorization or agenda for a given partner. As per our earlier discussion of Holmqvist’s (2004) categorization of organizational learning processes, this dimension also underlines a rich and promising area of inquiry.

Collaborative orientation depicts the nature and climate of a given collaboration: participatory, involuntary, and antagonistic. This is particularly important when looking at knowledge flows across partners. Under a participatory setting, one would expect the most favorable conditions for learning. The roles of the teaching partner and learning partner (Inkpen, 2001) are well specified, harmonized, and accepted. At the opposite, under an antagonistic setting, partners fight or co-habit at best. Learning is difficult. Partners adopt explicit protective measures, deploy shielding mechanisms, and engage in defensive actions to protect the transparency of their competencies, particularly when the embodied knowledge is explicit and held by only a few experts (Hamel, 1991; Inkpen and Beamish, 1997, Simonin, 1999a, 1999b). The last category, involuntary collaboration, depicts a situation where one organization may not even be aware of its role as a contributing partner. Then, learning is likely to be moderate and truncated through partial access only. As an extreme form of involuntary collaboration, reverse engineering and hiring of competitors’ talents open a window on learning. Assessing collaborative orientation will provide a reliable gauge on the boundaries of learning opportunities and challenges ahead.