Why Foresight Is Not 20/20 and What We Can Do About It: Preferential Attachment and Organizational Science

Growing interest in social networks research among organizational scholars has generated considerable understanding of how structural characteristics of networks can explain observed organizational behaviors and outcomes (Granovetter, 1985, 1992; Powell and Smith-Doerr, 1994; Ibarra, 1993). However, the evolutionary dynamics by which these network structures emerge and evolve are substantially less well-understood (Gulati & Gargiulo, 1999; Watts, 2004). As a result, network structural analysis has enabled organizational researchers to develop improved hindsight or explanatory potential regarding how structural antecedents impact behaviors once they have been observed; but has yielded less insights into how to predict or design future network structures and their behavioral consequences. An improved understanding of how network structures emerge and evolve over time would help sharpen organizations’ foresight with regard to the outcomes that these structures influence, and is therefore, a subject of considerable theoretical and practical significance to organizational researchers. Organizational science has turned to three basic natural science models of network emergence: the Erdos-Renyi (ER) model of random networks (1959, 1960), the Watts-Strogatz (WS) model of small worlds (1998), and the Barabasi-Albert (BA) model of scale-free network