Path Dependence and Path Creation: Alternative Theoretical and Methodological Perspectives on Strategy, Innovation and Entrepreneurship


Path dependence has been a topic of continuous and enduring interest in strategy research on capabilities and routines (e.g. Teece et al., 1997), in entrepreneurship and innovation research on technological design trajectories (e.g. Garud et al., 2002), and in institutional economics research on the persistence of institutions and governance structures (e.g. North, 1990). The recent growth in interest has been spurred by a resurgence of meta-theoretical reflections across these literatures on questions of agency and adaptability, alongside historical and institutional path dependencies (e.g. Garud and Karnøe, 2001). The answers to these questions come in various shapes and sizes, including emerging theories of institutional entrepreneurship and a move towards a focus on the ‘micro-foundations’ of capabilities in strategy research and of institutions in organizational theory.

Given this surge in research on path dependence, alongside the emerging issues addressed by research in multiple fields, we invited scholars with an interest in path dependence to debate its definition and application across the management and organizational literatures, and to progressively suggest theoretical clarifications and methodological recommendations that could help research move forward.

The result is a simultaneously provocative and inspiring debate, as captured in this issue's Point–Counterpoint. Although we do not wish to steal any of the thunder from the authors of the Point–Counterpoint, suffice it to say that they take substantively different perspectives on the definition and usefulness of the notion of path dependence. At the same time, they recommend drastically different methods for researching the creation and persistent influence of strategic, technological, or institutional paths over time.

In their Point article, Vergne and Durand (2010) take their inspiration from the mathematical roots of the notion of path dependence and, using formal logic, refine the construct definition of path dependence. They define it as a process in which initial conditions are followed by a series of contingent (chance) events whose influence on the path taken is larger than the initial conditions. When this happens, it changes the initial outcome distribution, with outcomes varying with time. Indeed, once a path has been contingently selected in this way, it may be self-reinforced (for example, because of increasing returns) and may be locked-in, as an important outcome, when there is no exogenous shock that unsettles the entire system. With this strict definition, Vergne and Durand are able to differentiate path dependence from various other constructs which, at an abstract level, all suggest that ‘history matters’. These constructs include well-known notions such as absorptive capacity and resource accumulation in strategy research, and structural inertia and imprinting in institutional and population ecology research. Their refined definition also sets clear parameters to researching path dependence. Specifically, they suggest that path dependence can only be directly evidenced in simulations and experiments given that it ‘is practically impossible to isolate structural causes when we observe a unique historical trajectory’ (Vergne and Durand, 2010, p. 736).

Garud, Kumaraswamy and Karnøe (2010) take the opposite position in their Counterpoint article. Rather than continuing with efforts to clarify and refine our understanding of path dependence, they suggest shifting to a different ontology, which they term as path creation. Whilst they do not deny the importance of emergent outcomes at a systems level that may have a hold over individual actors, their interest is in understanding how actors are not only embedded in, but also co-construct, such systems. Their position contrasts with Vergne and Durand (2010) in that they theorize directly about actors as part of emerging paths or systems rather than actors being placed within, and thus being confronted and constrained by, such paths or systems. Path creation is defined from a relational ontology that considers agency as part of nets of unfolding actions that emerge around topics and events, with actors being able to influence (but not determine) the processes that unfold. Specific attention is given in this perspective to the narratives and narrative ability of actors, which allows them to mobilize past memories and present experiences in order to imagine alternative future states and actions. Consistent with their ontology, Garud et al. (2010) challenge the use of those methods, such as simulations and experiments, as advocated by Vergne and Durand (2010). They argue that these methods will have little ecological validity and may restrict researchers to a focus on a select set of testable parameters. Instead, the suggestion is to study path creation and path dependence in the real world, combining an overall longitudinal research design with a focus on ‘real time’ observations (as events unfold) and narrative accounts of actors who imagine and iterate between the past, present, and future to create, shape or follow viable paths.

Together, these two contributions in this Point–Counterpoint help clarify and advance our theoretical understanding of path dependence and path creation, and their application in research on management and organizations. Vergne and Durand's (2010) efforts help us define the construct space of path dependence vis-à-vis related notions in strategy and institutional research. Garud et al. (2010) shift the paradigmatic frame which allows them to broaden our understanding of embedded agency, as embodied in their theorization of path creation. We hope you enjoy the debate that unfolds in the Point–Counterpoint and, as always, we encourage you to engage with it using the new Correspondence feature on the JMS website at http://www.respond2articles.com/jms. Please feel free to post comments or to elaborate on any of the arguments.

The Editors

Ancillary