Modeling interregional migration and spatial attractiveness: A gravitybased panel approach

Internal migration plays a pivotal role in shaping regional equity, labor market dynamics, and the spatial allocation of public services. This study employs an augmented gravity model to examine the determinants of interprovincial migration in Mongolia, utilizing balanced panel data spanning the period from 2000 to 2024. The model incorporates a multidimensional set of economic, spatial, social, and environmental variables and applies robust estimation techniques, including Estimated Generalized Least Squares (EGLS) and Seemingly Unrelated Regressions (SUR), to account for heteroskedasticity and cross-equation error correlations. Empirical results confirm that provinces with higher GDP per capita, elevated wage levels, and better access to healthcare attract significantly more immigrants. In contrast, greater distance from the capital and limited employment opportunities function as migration deterrents. Residual-based diagnostics further reveal spatial heterogeneity, indicating that certain provinces, including Khentii, consistently exceed predicted inflow levels, suggesting the influence of unmeasured factors such as regional identity, perceived safety, or cultural embeddedness. Conversely, provinces like Sukhbaatar exhibit persistent negative residuals, reflecting structural barriers, inadequate infrastructure, and limited access to essential services. A comparative assessment of the eastern provinces reinforces these findings, highlighting pronounced differences in regional migration appeal and underscoring the importance of localized development strategies. The study emphasizes that national migration policies must move beyond economic incentives, also to consider quality-of-life factors, ecological resilience, and social infrastructure gaps. While the analysis provides meaningful insights for spatial policy and regional development, it is limited by its reliance on aggregate data and the lack of household-level behavioral information. Future research should adopt mixed-methods approaches, integrating micro-level surveys and spatial econometric techniques to provide a more comprehensive model of migration decisions in transition economies. This study contributes to the migration literature by offering empirically grounded and policy-relevant evidence tailored to the unique context of a sparsely populated developing country.