Abstract It is increasingly common to study mobility and migration of individuals between social and physical locations as networks in which locations are nodes connected by mobile people. This conceptualization as mobility networks facilitates the analysis of how individuals influence one another in their mobility destinations. Technically, this amounts to analysing interdependence between individuals’ mobility paths. A recently proposed framework—the endogenous log-linear model (ELMo)—allows the statistical modelling of these social processes and, therefore, dependence in mobility, combining insights from exponential random graph models (ERGMs) and log-linear models. However, little attention was paid to how such models should be specified in a principled, theoretically informed way. In this study, we apply statistical theory to propose model specifications that can be used to analyse emergent structures in mobility. We first reformulate the model under analysis as a conditional multinomial logit with dependent observations. Subsequently, we show how to specify models that (i) are based on clear dependence assumptions on the individual level, that (ii) have a clear individual level interpretation, and that (iii) avoid (near-)degeneracy, a common problem for models with dependent observations. We end with an example application pertaining to the mobility of computer science faculty between university departments.
Hoffman et al. (Thu,) studied this question.