Working Paper: NBER ID: w22397
Authors: Dora L. Costa; Matthew E. Kahn; Christopher Roudiez; Sven Wilson
Abstract: At the end of the U.S Civil War, veterans had to choose whether to return to their prewar communities or move to new areas. The late 19th Century was a time of sharp urban growth as workers sought out the economic opportunities offered by cities. By estimating discrete choice migration models, we quantify the tradeoffs that veterans faced. Veterans were less likely to move far from their origin and avoided urban immigrant areas and high mortality risk areas. They also avoided areas that opposed the Civil War. Veterans were more likely to move to a neighborhood or a county where men from their same war company lived. This co-location evidence highlights the existence of persistent social networks. Such social networks had long-term consequences: veterans living close to war time friends enjoyed a longer life.
Keywords: Civil War veterans; social networks; migration; health outcomes
JEL Codes: J61; N91; R23
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Presence of an additional veteran from the same company (Y50) | Increased probability of living in that county (R23) |
| Presence of veterans from the same birth city (J45) | Increased probability of living in that county (R23) |
| Living near fellow company members (R23) | Reduction in probability of dying (J17) |
| Presence of social networks (Z13) | Migration decisions (F22) |
| Avoidance of areas with high immigrant populations (R23) | Reflection of veterans' social and ideological preferences (H56) |
| External factors like pension income (J26) | Affects locational choices (R30) |