Modes of interorganizational imitation: the effects of outcome salience and uncertainty

Administrative Science Quarterly,  Sept, 1997  by Pamela R. Haunschild,  Anne S. Miner

• E-mail
• Print
• Link

Many early theories emphasizing the importance of organizational context portrayed organizational environments as broad, static entities, characterized by stable properties such as turbulence or munificence (e.g., Lawrence and Lorsch, 1967). In the past two decades, however, theorists have increasingly emphasized processes through which individual organizations may be influenced by other organizations (Pfeffer and Salancik, 1978; DiMaggio and Powell, 1983). One such process is interorganizational imitation, which occurrs when one or more organizations' use of a practice increases the likelihood of that practice being used by other organizations. Some organizational research has explored specific mechanisms through which such imitation may unfold. DiMaggio and Powell (1983), for example, argued that the movement of professionals might lead organizations to imitate actions taken by other organizations, while Davis (1991) and Haunschild (1993) showed that interlocks may direct interorganizational imitation to particular firms.

In this study, we explore a different issue. We examine broad but distinct modes of selective imitation and predict conditions that will moderate their influence. Because different theories have implied several different potential imitation modes (e.g., DiMaggio and Powell, 1983; Levitt and March, 1988; Galaskiewicz and Wasserman, 1989), we first distinguish three fundamental bases for imitation. With frequency-based imitation, organizations execute practices previously used by large numbers of other organizations. With trait-based imitation, organizations use practices previously used by other organizations with certain traits, such as large size. With outcome-based imitation, organizations imitate practices that appear to have had good outcomes for other organizations in the past and avoid practices with bad outcomes.

Although it is likely that all three imitation modes occur, much organizational research has tended to emphasize one subset of modes over the others. Early neoinstitutional research, for example, emphasized frequency and trait imitation. Theorists argued that such imitation often arises from the pursuit of legitimacy or the widespread use of taken-for-granted practices (Tolbert and Zucker, 1983; DiMaggio and Powell, 1983). At the other extreme, much economic and technology research has emphasized efficient outcome imitation, assuming organizations imitate those practices that have clearly produced valuable economic returns for others (Griliches, 1957; Mansfield, 1961; Reinganum, 1981). Between these poles, descriptions of the diffusion of innovations (Rogers, 1995) and interorganizational learning (Levinthal and March, 1993) have emphasized both social and technical imitation modes. Thus, while several literatures point to the existence of the three imitation modes, we lack systematic, empirical, organization-level research that examines all three modes in a single setting, along with factors that affect their relative influence. There are few quantitative empirical studies of interorganizational learning, and none that examine all three imitation modes. The diffusion of innovation literature contains many empirical studies, but they are primarily at the individual level of analysis (Rogers, 1995) and rarely examine the effect of all three imitation modes at the same time. Finally, while all four literatures -- neoinstitutional, economic, learning, and diffusion of innovations -- often highlight factors that might influence the relative strength of imitation modes, they contain few formal tests of potential moderators.

Our first goal, then, was to test whether all three selective imitation modes operate in a group of organizations. Our second goal was to examine the relative impact of two factors widely theorized to influence imitation: the salience of outcomes experienced by other organizations and contextual uncertainty. Some communication and organizational learning theories, for example, imply that organizations will be more likely to imitate practices that produce highly salient, attractive outcomes (March, Sproull, and Tamuz, 1991). Neoinstitutional and social information theories imply that uncertainty will shift the balance between imitation modes (Pfeffer, Salancik, and Leblebici, 1976; DiMaggio and Powell, 1983). Specifically, greater uncertainty should enhance the impact of frequency and trait imitation but reduce or leave unaffected outcome imitation.

We explore these questions by examining whether the facts about other firms' use of a particular investment banking firm later affect whether a focal firm chooses that investment banker in a set of 539 acquisitions from 1988 to 1993. Interorganizational imitation is defined here as occurring when one or more organizations' execution of a practice increases the likelihood of that practice being used by others. In this framework, imitation is not effective or ineffective, conscious or unconscious, rational or irrational. it simply represents a way in which the actions of one group of organizations affects the actions of others.