The Effect of Military Campaigns on Political Identity: Evidence from Sherman's March
This repository contains files neccessary for replicating figures and tables in "The Effect of Military Campaigns on Political Identity: Evidence from Sherman's March" (pdf available here).
Note that 1880 full count US census data and 5% samples of 1930 census data need to be downloaded from IPUMS for full replication.
Abstract
I use the military march of Union general William Sherman during the American Civil War to estimate the effects of wartime violence and destruction on post-war voting behavior and personal identity. First, I examine how the march influenced the support for the Democrats throughout the 19th and 20th centuries. Second, to proxy for the strength of Southern identity, I construct several variables from both historical and contemporary sources. These variables include the share of individuals likely named after famous Confederate generals, the relative frequency of streets likely named after Confederate figures, and the presence of Confederate monuments. The results show mostly small and statistically insignificant effects of the march on Democratic vote share. For some outcomes proxying for Southern identity, I find a significant positive effect; however, these results are not robust across different model specifications. Overall, the results suggest that Sherman’s march did not have a transformative impact on the politics and personal identity in the US South.