When trust fades... : can optimal mechanisms for policy decisions always be designed?

Governments must usually take policy decisions with an imperfect knowledge of the economic actors' type or the actors' effort level. These issues are addressed within the framework of classic adverse selection or moral hazard models. I discuss in this paper how would the government’s and the economic actors’ behavior change if relevant information is double asymmetric, that is, it is not just the government that has limited information about the agents’ type or effort level, but the economic actors also lack perfect information about the government's trustworthiness. Using the modeling tools of mechanism design I prove in the paper, that government - as principal - is only capable of applying "perverse" incentives towards the economic agents: it punishes well-behaving agents while it rewards the badly behaving ones. I apply the theoretical models to the regulatory issues of network industries, and specifically to the ICT industry.