Organizational Labor Flow Networks and Career Forecasting.

The movement of employees within an organization is a research area of great relevance in a variety of fields such as economics, management science, and operations research, among others. In econophysics, however, only a few initial incursions have been made into this problem. In this paper, based on an approach inspired by the concept of labor flow networks which capture the movement of workers among firms of entire national economies, we construct empirically calibrated high-resolution networks of internal labor markets with nodes and links defined on the basis of different descriptions of job positions, such as operating units or occupational codes. The model is constructed and tested for a dataset from a large U.S. government organization. Using two versions of Markov processes, one without and another with limited memory, we show that our network descriptions of internal labor markets have strong predictive power. Among the most relevant findings, we observe that the created by our method based on operational units possess a power law feature consistent with the distribution of firm sizes in an economy. This signals the surprising and important result that this regularity is pervasive across the landscape of economic entities. We expect our work to provide a novel approach to study careers and help connect the different disciplines that currently study them.