Many have argued that the main function of organizational design is to manage and direct the time, attention, and flow of information among individuals and organizational units (Cyert and March, 1963; March and Simon, 1958; Ocasio, 1997). This function has become increasingly challenging over time, as the dynamics of globalization have exerted simultaneous pressures on organizations to be more efficient and competitive, while at the same time increasing the need for learning so that organizations and the individuals within them can keep up with new technologies and demands. Though conceptually the time and energy put toward learning and skill development should foster improved performance, some analyses suggest that learning and productivity can work at odds with one another (Bunderson and Sutcliffe, 2003; Edmondson and Singer, 2008; Ren, Carley, and Argote 2006). To accommodate the demands for higher productivity and faster learning, organizations have increasingly turned to using smaller and more flexible work units, such as teams, to accomplish their most important tasks. Over time the use of teams has shifted from team-based work in hierarchical structures, to team-based work in matrix structures, and ultimately to team-based work in multi-team systems (Hatch and Cunliffe, 2006; Hobday, 2000; Malone, 2004; Marks, Dechurch, Mathieu, Panzer, and Alonso, 2005; Scott and Davis, 2006) in part to enable the knowledge and skills of individuals and smaller units to be leveraged across more projects in more parts of the organization.
One increasingly common design choice is to assign individuals to multiple teams simultaneously. Some surveys (for example, Lu, Wynn, Chudoba, and Watson-Manheim, 2003; Martin and Bal, 2006; O’Leary, Mortensen, and Woolley, 2011) estimate that simultaneous membership on more than one team appears to be the norm for at least sixty-five percent of knowledge workers across a wide range of industries and occupations in the US and Europe (Zika-Viktorsson, Sundstrom, and Engwall, 2006). Some surveys place the percentage of knowledge workers who are members of more than one team as high as 94.9 percent (Martin and Bal, 2006) and in at least one company (Intel) twenty-eight percent are on five or more (Lu et al., 2003). Multiple team membership seems especially common in many industries and settings in which learning and productivity are both especially critical, including information technology (Baschab and Piot, 2007), software development (Shore and Warden, 2007), new product development (Edmondson and Nembhard, 2009; Wheelwright and Clark, 1992), some consulting firms (Milgrom and Roberts, 1992), and education (Jones, 1990), but appears to be widespread in a variety of other contexts such as auto repair (Madono, 1998) and healthcare (Richter, Scully, and West, 2005).
As more teams share members, there is increased resource interdependence among different units of the organization. Researchers have begun to refer to this as a ‘multi-coupled project organization’ (Söderlund, 2002: 428) or as an instance of intra-organizational connectivity (Lazer and Friedman, 2007), or simply as multiple team membership (MTM) (Mortensen, Woolley, and O’Leary, 2007). Intra-organizational connectivity has positive effects on organizational learning. As noted by Kang, Morris, and Snell (2007), to understand organizational learning, it is important to consider the pattern of relationships among parties within a firm (Burt, 1992; Coleman, 1988; Uzzi, 1997). As intra-organizational connectivity increases, organizations have more paths along which information can flow, which increases the likelihood that any two potentially complementary pieces of information will be brought together and simultaneously decreases the likelihood that any potentially valuable piece of information is stuck in one part of the organization and ‘lost.’ The more often employees interact, the more opportunities they have to identify and utilize idiosyncratic knowledge (Hansen, 1999; Krackhardt, 1992; Nelson, 1989; Uzzi, 1997).
High interactivity across teams, such as that arising from shared membership, results in more integrated knowledge across those teams (Newell, Goussevskaia, Swan, Bresnen, and Obembe, 2007). Intra-organizational connectivity creates built-in boundary spanning capabilities across teams and improves information sharing in the organization (Ancona and Caldwell, 1992; Hansen, 1999; Lazer and Friedman, 2007). High levels of multiple team membership create embedded Simmelian ties, which are much stronger predictors of innovation than weak ties (Tortoriello and Krackhardt, 2010). People will carry lessons learned across units, managers at higher levels will have more sources of information about various projects and their staff (Meyer, 1994), and more opportunities will exist for the propagation of ideas across the organization (Subramaniam and Youndt, 2005). Nobeoka, Cusumano, and others (Cusumano and Selby, 1995; Nobeoka, 1995) have shown how intra-organizational connectivity enhances organizational learning via enhanced cross-project learning. Such organizational learning by working across projects may break up ‘collaborative dead-ends’ more than simple interaction across team boundaries would (Dornisch, 2002) and improve organizational learning (Carlile, 2004; Hansen, 1999; Lazer and Friedman, 2007; Marrone, Tesluk, and Carson, 2007).
As teams within an organization become increasingly interconnected via shared members, organizations are more able to shift individuals fluidly and quickly from team to team to react to changing environmental conditions. In particular, this allows organizations to leverage their dispersed resources in multiple teams without incurring the costs typically brought about by restructuring or reassignment, allowing organizations to accomplish more with a given set of resources. This argument is consistent with Lojeski et al. (2007) who found that higher organization-wide multi-tasking is a key contributor to organizational productivity in the last decade.
Despite its advantages, it is possible that organizations can become too interconnected. Organizational slack refers to the supply of uncommitted resources in the organization (Bourgeois, 1981; Cyert and March, 1963). Slack resources allow organizations to respond to environmental events by allowing time to experiment and reflect on their responses (Meyer, 1982). Even in the absence of exogenous shocks, organizational slack can enable management to experiment with new postures in relation to the environment through innovations in new product development (Nohria and Gulati, 1996; Tushman and O’Reilly III, 1996) or in management style (Bourgeois, 1981). Such reflection and experimentation is critical for organizational learning. Thus, a reduction in organizational slack, which can occur by committing members to too many teams simultaneously, limits organizational learning.
In this chapter, we take a closer look at how team learning, both within and across organizational teams, serves as a fundamental building block of organizational learning. We start with a review of the literature on team learning, which has emphasized studies of learning within teams. We then draw connections between this research and a new research agenda to examine team learning across teams, with a particular focus on multiple team membership. Finally, we discuss the practical implications of how the tension between productivity and learning shapes organizational learning strategies in modern, global companies.