Migration and the Geography of Life Expectancy: Selective Sorting and Spatial Dependence Across 172 Countries

International migration has grown rapidly in recent decades, yet its relationship with population health is difficult to disentangle from the economic, structural and geographic characteristics of the places migrants move to. This study examines whether international migration is associated with population‐level life expectancy after accounting for confounding and the data's spatial structure. Data for 172 countries (189 in the spatial analyses) for 1990–2024 were assembled, combining migrant stock statistics from UN DESA, life expectancy estimates from Gapminder, and economic indicators from the World Bank. The analysis used cross‐sectional and panel data methods (pooled Ordinary Least Squares, country fixed effects, and first differences), spatial lag and spatial error models, spatial panel models, multiscale geographically weighted regression (MGWR), and an origin–destination analysis of bilateral migrant stocks. Migration and life expectancy were positively associated across countries (r = 0.42, p < 0.001), but the cross‐sectional migration coefficient was reduced by 47% after adjustment for GDP per capita, and no association was detectable within countries over time (fixed effects β = −0.13, p = 0.051; first differences β = −0.04, p = 0.744). Life expectancy was strongly spatially clustered (Moran's I = 0.68), and migration showed no detectable association with life expectancy once spatial dependence was modelled. MGWR indicated that the migration–life expectancy relationship was effectively global in scale and non‐significant, whereas unexplained place effects varied at a sub‐continental scale. Eighty percent of international migrants lived in a country with higher life expectancy than their country of origin. These findings suggest that the cross‐national association reflects the selective sorting of migrants towards already healthier places, compounded by spatial dependence, rather than an independent effect of migration on population health. More broadly, they extend the selective‐sorting framework from the subnational to the international scale and highlight the importance of geographic structure in interpreting migration–health relationships.