The Leninist Vanguard's Dilemma: Information and Coordination in Revolutions

We model the role of a Leninist vanguard as a sequential coordination game in which the vanguard and a follower must decide whether to revolt to secure an uncertain prize. The true value of the prize is unknown, but each agent privately observes a binary signal about its worth. The vanguard must act first and publicly, absorbing the strategic risk of punishment if isolated. The follower then decides whether to join, weighing the vanguard's costly action against her own private information. We characterize the generically unique Perfect Bayesian Equilibrium surviving Forward Induction. The parameter space structurally partitions into four regions based on the relative precision of the agents' signals and on their joint precision. We show that through the application of Forward Induction, these four regions collapse into three distinct equilibrium outcomes. When the vanguard is better informed about the prize, the equilibrium is first-best efficient. However, when the follower is better informed, coordination partially or entirely breaks down. If their joint precision is too low, we get to total inactivity. If joint precision is high enough, a separating equilibrium exists but remains strictly inefficient. Ultimately, improving the follower's information about the uncertain prize paradoxically widens the efficiency gap, while improving the vanguard's precision can restore the first-best outcome. These results show that the role of a Leninist vanguard is effective if and only if it is better informed than the followers.