Do dictators signal strength with electoral fraud?

What role does electoral fraud play in nondemocracies? In this paper, we offer an empirical test of a popular idea that authoritarian governments use elections to engineer overwhelming victories with electoral fraud thus deterring potential opposition from challenging the regime. Using the data from the Russian Parliamentary elections in 2011 and a regionally representative public opinion survey, we find that the geographical allocation of electoral manipulation was the opposite of what the theory would imply: more manipulation happened in the areas where the regime was more popular. We also find that higher margins of victory for a pro-regime party failed to deter subsequent mass protests. We argue that these empirical patterns could be better explained by other mechanisms, such as Bayesian persuasion, efficient allocation, and information gathering