Climate and nomadic migration in a nonlinear world: Evidence of the historical China

This paper deals with climate change and nomadic migration relationships at a long-term and wide geographical scale using a statistical approach in the vein of Bai and Kung (Rev Econ Stat 93:970–981 2011). More precisely, it presents a reassessment of these relationships in a nonlinear world using threshold regressions, time varying-copula, and nonlinear causality tests. The large amount of historical records in China enables us to re-interpret the link between climate and historical social dynamics (Hsiang et al., Science 341:1235367 2013) through different regimes of temperature and precipitation. Our nonlinear results have confirmed the role of climatic factors on migration. However, we find that climatic factors have affected the migration flows only during some sub-periods (especially 300–500 AD and 1050–1300 AD). As a consequence, the importance of the precipitation effect on migration has probably been slightly overestimated whereas the role of temperature anomalies has been underestimated. Copyright © 2020 Springer Nature B.V.