In the mechanistic perspective firm strategy, the environment, and the firm’s stock of resources, structure, and work flow technology are often treated as given discrete categories or states that coalesce to create static efficiency (e.g., economies of scale), fit, and configuration (e.g., Galunic and Eisenhardt, 1994). Strategic management is viewed as a one-time sequence of formulating and implementing a single choice rather than a continuous process. Strategy-making mechanisms are assumed to be in place, and learning, history, and processes are downplayed. Strategy formulation and implementation activities are condensed in time and their duration is inconsequential. The choice part of the model often involves once-and-for-all choices for which past and subsequent choices are not considered, and where there is no distinction between initiation of a new alternative and the continuation of an existing one (March and Simon, 1958).
The discrete view of time is also evident in the research models being used. Most mechanistic studies use variance models, cross-sectional in design. Variance models are concerned mainly with what the relative explanatory power of different determinants of abstract entities (i.e., strategy and performance) are rather than how these entities are formed (Mohr, 1982). Although process explanations featuring the role of history and learning were central in the founding of the main theories (e.g., Selznick, 1957; Penrose, 1959; Chandler, 1962), they have been largely neglected by subsequent research.
Underlying this particular view of time and the focus on variance models is the idea of efficient historical process —an evolutionary process that moves rapidly to a unique steady-state equilibrium solution, conditional on current environment conditions, and thus independent of the historical path. A static alignment at a given time is the product of a rapid optimizing process. The process is assumed to lead to improvement of fit and ultimately to the one most suitable (March, 1994).
In early versions of the main underlying research programs causality ran from environment to strategy and performance (in the SCP model), from strategy to structure (and performance) in the SSP model, and from resources to strategy and performance in the RBV model. Moreover, as postures, strategies are mainly responses to given constraints rather than means to influence them or create new environments (Porter, 1980). Additionally, at its core, positioning analysis often assumes no responses from competitors and other players (Ghemawat, 1991), and value chain analysis largely represents sequential interdependence (Stabell and Fjeldstad, 1998). The sequential flow of the design model and the view of strategic choice also illustrate directionality. Feedback loops are either implicit, as from implementation to formulation, or absent, as in the case of performance influences on other elements. Choice constitutes a constrained optimization problem where the choice set is exogenous and given (Porter, 1991).
Because of their disciplinary and historical roots, the main models have been developed from the ground up as fragmented middle-range theories rather than as lower-level theories stemming from an integrated overview of strategy. Moreover, each of the research programs has focused on a different element of the strategy picture: environment, resources, and organizational structure. This division of labor between programs of research has facilitated scientific progress—but at a price (Schendel, 1994). The SCP and the industry structure model have been criticized for lacking a theory of the firm’s organization (Teece, 1984) and as generally ignoring the inner context of strategy (Pettigrew, 1987). The SSP has been criticized for not paying attention to competition (Galbraith and Nathanson, 1978), and research in RBV has only recently begun to explore the mutual dependence of internal resources and competition (Levinthal and Myatt, 1994; Priem and Butler, 2001). By and large, theoretical differentiation has considerably hampered the recognition of multiple and reciprocal causality between these distinct elements (Henderson and Mitchell, 1997).
Уважаемый посетитель!
Чтобы распечатать файл, скачайте его (в формате Word).
Ссылка на скачивание - внизу страницы.