In this chapter, I address the issue of selection bias more directly. First, I present a comparative case study using most-similar research design between the two similar princely states of Awadh and Hyderabad, which shows that historical contingency determined by external geo-political circumstances prevented British from being selective and led to direct rule in Awadh vs. indirect rule in Hyderabad. Second, I develop a new instrument for the British choice of indirect rule through princely states based on the exogenous effect of major European Great Power wars that decreased the ability of the British to fight wars of annexation to bring additional territory into direct rule and increased their tendency to sign treaties of indirect rule with Indian states on the frontiers of British direct rule. The instrumental variable (IV-2SLS) analysis is a major empirical contribution and allows an estimate of the causal effects of colonial indirect rule on Maoist insurgency. I also develop a fine-grained typology of different types of princely states and show that warrior states like Mysore had higher development, while successor states like Hyderabad and feudatory states like Bastar had more inequality and less development and thus Maoist insurgency.

I use archival and fieldwork-based qualitative data to do process tracing of the causal mechanisms of the crucial pathway case of the Maoists in Chhattisgarh. The northern and southern parts of Chhattisgarh had colonial indirect rule through feudatory princely states, which created weak state capacity and despotic extraction of land revenue and natural resources through landlords and feudatory chiefs. This created tribal grievances that persisted in the postcolonial period in the 1950s-80s through path dependence of these mechanisms, which were mobilized by the People’s War Group (PWG) Maoists, leading to high levels of Maoist rebel control by the 1990s. In contrast the central districts of Raipur and Bilaspur had been under British direct rule and had relatively higher levels of development and less exploitation of forest and natural resources of tribals, and so the Maoists did not succeed there. I describe the history of the PWG Maoists and how they contributed to the welfare of the tribals and opposed natural/forest resource exploitation, which highlights the role of rebel agency. Finally, I develop a novel constituency-level dataset to test the theory and show that former princely state constituencies had more Maoist control.