Moving From a Sketch to a Painting: Toward an Informative Mobility Effect

Abstract

This article develops an identification strategy for estimating the mobility effect in linear models by enriching the mobility term with theoretically grounded proxy measures that do not linearly covary with class origin and destination. It introduces three substantively informative metrics of social mobility: permeability, defined as the difference in frequency or probability mass between mobility positions and their corresponding nonmobile diagonals; atypicality, captured by the deviation between individuals’ predicted probabilities of occupying a given mobility position and those associated with the diagonal; and social distance, measured as the covariate-based Gower distance between mobility positions and their diagonal counterparts. By incorporating these proxies, the proposed framework enables simultaneous estimation of the net effects of origin, destination, and mobility, while relaxing restrictive functional constraints in conventional specifications. This reformulation effectively addresses previously underexplored research scenarios, including reference-group shifts, counterfactual decomposition, sociodemographic matching, and continuous measures of socioeconomic status. Two empirical examples are provided to demonstrate the analytical leverage and interpretive gains of the proposed approach.