Whip or carrot? effect of socio-economic reforms on violence

Abstract

Can socio-economic government schemes help reduce violence? The existing lit erature reveals a lack of consensus on the direction of effect that welfare schemes have on violence. In this paper, I attempt to look at the umbrella Security Related Expenditure (SRE) scheme implemented by the central government to address left wing extremism in India. The decision to treat any geographical area with welfare policies is often guided by the patterns in violence itself. This results in an en tangled relationship of welfare schemes affecting violence and violence affecting the probability of receiving the schemes in the first place. In this regard, I implement a spatial regression discontinuity design to isolate the effect of the SRE schemes announced in 2018 on violence. A subdistrict level panel dataset was created for this purpose using fuzzy matching algorithms and geospatial techniques to maximize replicability. I find that violence post 2018 is lower by approximately 4 events in Chhatisgarh and 6 events in Odisha for the ‘just’ treated subdistricts as compared to those which were ‘just not’ treated. Statistically significant results are observed only for these two most affected states. A placebo test using violence before reforms shows that the discontinuity did not exist in the absence of the 2018 schemes.