Between the Devil & the Deep Sea:Civilian Perception of India's Naxal Insurgency

• In Event: Capitalism and Development in the Global South

Abstract

What explains variation in civilians’ attitudes towards insurgent groups? Studies of civil war that have theorized civilian loyalties to either the insurgent or the state have done so based on observable behaviors and actions instead of relying on hard-to-infer attitudes in conflict zones. This is not surprising due to paucity of attitudinal data from conflict zones. More importantly, while some existing frameworks leave open the possibility of neutrality, they assume that civilians are aware of insurgent presence. In response, we posit a typology of civilian attitudes during wartime and visualize them as lying on a continuum from full support for the counterinsurgent to full support for the insurgent. This typology accounts for the possibility of both, neutrality and lack of knowledge about insurgents among civilians. Empirically, we leverage a novel public opinion dataset (survey conducted in August 2010; n = 3990) from a random sample of the civilian population in India’s Naxal conflict zones. We show that civilians exhibit interesting variation in both, their level of knowledge and support of insurgents. Further, we evaluate the determinants of such variation in attitudes and argue that the 'control-collaboration' model holds true with regards to individual attitudes in conflict zones. Specifically, an increase in civilians’ perception of a political actor’s level of control increases support in favor of it. However, violence against civilians in a district in preceding years reduces the rate of favorable response in support of Naxals. At the same time, conventional explanations – such as the presence of the state’s infrastructural apparatus, level of media exposure, support for governmental policies and demographic characteristics such as economic class, caste, gender and region are unable to explain the variation in attitudes. This null-result on demographic characteristics is especially important given the scholarly consensus that the Naxal conflict is a class-based movement. It is also timely in its response to a new body of work that connects Naxal violence to adverse income shocks.
Given that such survey-based attitude measures are susceptible to falsification in conflict zones where survival considerations dominate, we confirm the robustness of our findings through two means. First, we compare the results from our public opinion dataset that was exclusively conducted in districts affected by Naxal insurgency in 2010 with another all-India, post-election survey conducted after the 2009 general elections, where we obtain very similar results. More interestingly, we analyze enumerators’ post-survey evaluation of situational and body language cues at the level of each individual respondent. Here we find that the significance of our results remains unchanged even when we exclude samples of the population whose ability to convey true preferences was judged by survey enumerators to have been compromised.

Authors

• Dipin Kaur, Yale University

• Rahul Verma, University of California, Berkeley