Working paper • Vol/Iss. . • Harvard Scholar • • Published In • Pages: i-64 •
All analyses are repeated using an alternative measure of conflict (UCDP GED data). Results from the UCDP GED data and ACLED data are very similar. Additionally, the authors account for unobservables using Spatial RD Estimates, and find no evidence for omitted factors generating biased estimates.
Documents and Hypotheses Filed By: jack.dunnington emily.pitek