Quantifying the Power of Social Connectedness: Evidence from the Great Migration

Jingyi Liu, University of Purdue

In this paper, I estimate the causal effect of hometown social connectedness on life outcomes of black migrants, using the Great Migration as a natural experiment. To do so, I use census tree data spanning 1910-1940 to construct a historical migration network that captures migration pattern during the First Great Migration. I then construct the hometown migration network exposure in 1940s in each southern county using linked census from MLP project and instrument the social connection exposure by World War II investment shocks. Given the exogeneity of the World War II shocks and pre-existing historical network, I estimate the causal effect of social connectedness on life outcomes of southern black migrants who migrated from south to north between 1940 and 1950. This is an ongoing project. Preliminary finding confirms two historical narratives. Firstly, compared to the First Great Migration, the Second Great Migration led to more concentrated destination choices. During the first great migration, migrants dispersed among industrial cities in the North, while in 1940s, migrants clustered in cities where WWII jobs were highly concentrated. Secondly, compared to the First Great Migration, migrants in 1940s redirected from the Northeast to the West and some new industrial cities in the Midwest. The next step of this project is to explore the causal effect of historical migration network using changes in migration pattern as the source of variation. The causal inference is expected to be done within the next two months.

No extended abstract or paper available

 Presented in Session 212. Using IPUMS Full Count Data: Spatial Dimensions in Historical Data