Human capital shock and economic development: Evidence from Partition of India

This paper studies the socio-economic effects of skilled refugees inflows under limited state capacity using the 1947 Partition of India as a historical natural experiment. The Partition triggered one of the largest forced migrations in his- tory, displacing approximately 15-18 million people within a few years. Refugees were culturally and linguistically similar to natives but significantly more liter- ate, creating a positive human capital shock. Using a newly constructed village- level panel dataset spanning six decades, I examine how this sudden influx of skilled refugees shaped long-run economic outcomes. To address potential endo- geneity in settlement patterns, I employ an instrumental variable strategy that exploits variation in distance from village centers to the India–Pakistan border. The results reveal large and persistent gains in literacy lasting several decades, with particularly pronounced effects for women. In contrast, local employment structures adjust slowly, with limited evidence of rapid structural transformation. Despite this, refugee-exposed villages experience higher consumption levels and higher incomes. The effects are especially strong in districts historically linked to canal colonies, suggesting that the interaction between refugee inflows and skill composition of migrants together played an important role in shaping long-run outcomes.