Centralized state response is almost universally considered as the first best option for managing pandemics. This paper argues that in reality, states fall short of this ideal. Knowledge problems hinder the government's disease acknowledgement, resource allocation and feedback, while compulsion, political predation, and corruption exacerbate the situation further. On the other hand, a decentralized, voluntary response may overcome knowledge problems and opportunism through effective signals such as prices, better incentives, local knowledge, and internalization of disease-related externalities. This results in a speedier, more effective, and responsive pandemic management, which also accounts for differences in risk preferences, and becomes a feasible second-best option. I then apply these insights to Russia and its history of disease management. I focus on three diseases: plague, cholera, and COVID-19, which results in a historical analysis that spans multiple centuries and different institutional settings. I show that government-led pandemic response in Russia has been riddled with knowledge problems, abuse, political predation, and violence. On the other hand, a decentralized response was quicker and more effective at managing the disease, often emerging even in the most unfavorable circumstances.