Organizational Research Methods, Ahead of Print.
We present a dynamic multilevel framework for analyzing the mutual dependence of change in interorganizational networks and internal organizational structure. Change occurring at the former (interorganizational) level involves decisions to change the portfolio of network ties to external partners. Change occurring in the latter (intraorganizational) level involves decisions to change the portfolio of internal activities. We estimate a recently derived class of stochastic actor-oriented models (SAOMs) that we adopt and adapt to specify how decisions to change internal portfolios of activities and external portfolios of partners are connected by theoretically derived multilevel mechanisms that link organizational and network structures. We show that statistical models for multilevel networks reproduce with high fidelity the structural regularities observed in the distribution of (a) activities within organizations, (b) network ties between organizations, and (c) knowledge available in the organizational field. We discuss the implications of the study for theory development, and for empirical research on interorganizational and other kinds of multilevel networks.