Year: 2025
Type: preprint
Abstract: Theoretical and empirical research on migration has long been divided between studies of migration drivers and those focused on spatial patterns. This divide has limited methodological development: existing model evaluation strategies rely on goodness-of-fit statistics that overlook the spatial structure of migration. This paper introduces a simulation-based procedure that addresses this gap by evaluating how well explanatory models—and the theories they operationalize—reproduce observed spatial patterns. The procedure uses empirical models to generate theory-driven synthetic migration systems, which are then compared to real-world data. The procedure is flexible and scalable, suitable for both internal and international migration settings where origin-to-destination moves are recorded. It offers two key advantages: first, it provides a robust spatial test of model accuracy; second, it yields interpretable counterfactuals for scenario exploration. To illustrate the method, I apply it to migration between 31 European countries from 2002 to 2021. Migration systems theory (MST) accurately reproduces three measures of spatial concentration in most years, while the gravity model consistently underestimates them. However, neither model accounts for the high reciprocity of flows nor correctly predicts major corridors in Northern Europe. These results underscore the need for spatially explicit model evaluation and demonstrate the value of the proposed approach.
Author: Micol Matilde Morellini
Institution: University of Oxford
Language: English
Topic: Migration and Labor Dynamics
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: gold
Funders: National Science Foundation, International Max Planck Research School for Environmental, Cellular and Molecular Microbiology, University of Oxford, Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung, International Max Planck Research School for Advanced Methods in Process and Systems Engineering, Economic and Social Research Council, Nuffield College, University of Oxford
OpenAlex ID: W4411959609