SLIDE 47 Introduction Theoretical Framework Empirical Framework Results and Discussion
Limitations
◮ Data quality
◮ ACLED is widely regarded as a useful and reliable source of
data for conflict studies (Raleigh et al. 2010; Weidman 2014; Maystadt et al. 2014)
◮ I cannot rule out the possibility that these results are due
to some characteristic of the data
◮ Next step: Perform the same analysis using different data
sources, such as UCDP/PRIO and SCAD
◮ Identification strategy
◮ It is less than clear if the necessary conditions exist for the
identification assumptions used in this analysis
◮ Conflict trends are not strictly speaking parallel, but are
“similar” in a statistical sense
◮ Next step: Perform additional robustness checks to control
for the potential impact of the ongoing trend in conflict events
Jeffrey R. Bloem University of Minnesota Good Intentions Gone Bad