The drowning-out effect: voter turnout, uncertainty, and protests

Conventional wisdom suggests that high turnout in a free and fair election is desirable; it might signify proper representation, help the government identify and address social discontent, and hence reduce the likelihood of protests. In this paper, I demonstrate that this intuition does not necessarily hold. By extending a bargaining theory, I hypothesize that high turnout “drowns out” the voices of dissenters, creates uncertainties over the social discontent, and thus causes post-election protests. I test this hypothesis using election-day rainfall deviation as an instrumental variable for turnout and apply it to a new constituency-level dataset of Indian elections. I also extend a new design-based method, called near-far matching, that makes the causal comparison more powerful, robust, and explicit. The result shows that higher turnout increases the occurrence of protests. This finding implies that electoral democracy can be inherently imperfect as a conflict resolution mechanism.