Unequal Land, Unequal Politics: Elite Persistence in India

Why does political inequality persist in unequal societies, where elections are often dominated by elite parties and candidates? Conventional models of political economy predict that higher inequality should be associated with higher levels of support for redistribution and left-wing parties. When these models have failed to explain voting outcomes, scholars have typically emphasized demand-side factors such as ethnic, social and cultural cleavages. Building on the theoretical foundation laid by Acemoglu and Robinson (2008), I argue that supply-side factors may be just as important, particularly in new democracies. I suggest that socioeconomic inequality may limit the supply of parties and candidates which represent non-elite groups. Elites in societies with higher inequality have greater incentive to invest in the preservation of existing structures. This investment can take several different forms, from the funding of elite parties to violence against members of subaltern groups. Such investment increases the costs of participation for non-elites, hence entrenching elite domination of politics. I evaluate these predictions by looking at quasi-random differences in colonial land tenure systems in India, building on the work of Banerjee and Iyer (2005). While Banerjee and Iyer (2005) posit that more unequal land tenure systems could have led to greater support for redistribution, I hypothesize that the opposite may be true. More unequal systems may actually produce greater elite domination of politics, manifesting in higher vote share for right-wing parties, lower vote share for subaltern-led parties and reduced subaltern representation across parties. Empirically, I test these predictions using an instrumental variable approach and a geographic regression discontinuity. Firstly, I adopt the instrument used by Banerjee and Iyer (2005), which uses the date of British conquest as an instrument for the colonial land tenure system in a particular district. This design assumes that the date of annexation, conditional on years under British rule, affected political outcomes only through their impact on land tenure. Secondly, I apply a geographic regression discontinuity (GRD) in Uttar Pradesh and Tamil Nadu, where land tenure borders cut across modern state lines. I find that areas with more unequal land systems were more likely to have higher levels of support for right-wing and elite-led parties. The effect is persistent across time and robust to different conceptions of political parties. I also provide some suggestive evidence for the aforementioned mechanisms, using data on the wealth of legislators.
The results have significant implications for our understanding of historical political development, representation, inequality and redistribution.