Abstract for:Relating Policy Differences to Behavior Differences across Family Members

Forrester’s (1973) statement set out a particularly lofty goal: put simply, our models should enable us to explain the different outcomes of systems by differences in the policies pursued by their leaders. However, most work has focused on establishing confidence in a model’s ability to explain the behavior of a single observed outcome (Barlas, 1989, 1994; Barlas & Carpenter, 1990; Homer, 1983; Oliva, 2004; Sterman, 2000). A few models, notably diffusion models and commodity models, have been shown to replicate the varied behavior of a larger family of examples. However, differences in behavior in those cases originate primarily from differences in model parameters unrelated to managerial policies (Homer, 1987; Meadows, 1970; Sterman, 2000). Studies that have related differences in enduring policies to differences in behavior are almost exclusively experimental studies (Gary, Wood, & Pillinger, 2012; Moxnes, 1998; Paich & Sterman, 1993; Sterman, 1989a, b). Our research links differences in managerial policies to differences in behavior in a field setting. We combine an established model of a single organization with a large data set of similar organizations. Our hope is to establish precedents for similar field studies, helping raise the influence of systems models in managerial practice, public policy, and education.