Optimal mechanisms against corruption: incentives, self reporting and delegation.

This Ph.D. thesis is a collection of four papers, each one corresponding to a chapter. The background challenge, common to all chapters, is the design of anti-corruption policies in a principal/supervisor/agent framework, where the supervisor and the agent can collude at the expense of the principal. The first two works tackle the issue of corruption from the perspective of law enforcement. The third and fourth chapters consider the problem of collusion from a mechanism design point of view. Accordingly, the focus is shifted from incentives to organizational responses, where the choice between centralized and decentralized organization of economic activity is relevant. More precisely, the first chapter studies the optimal compensation policy for a corruptible inspector, in charge with monitoring evasion from taxpayer. Namely, I discuss how the optimal compensation policy varies according to the timing of collusion, which is allowed to occur either before or after inspection takes place. Results show that increasing the inspector's bonus rate is a better policy than increasing the penalty rate, when corruption occurs after inspection. The contrary is true when the collusive agreement is established before the inspection. Implications for privatization of law enforcement and economic development are also analyzed. The second chapter analyses the impact of self reporting on law enforcement when officers are corruptible. The threat of corruption highlights two additional advantages to the use of self reporting beyond those identified elsewhere. First, by allowing individuals to self report their unlawful acts, the government is able to increase welfare by eliminating rents to its officers. Second, the introduction of self reporting further bene?ts those governments, which in its absence would tolerate corruption, by allowing them to fully eliminate corruption. The third chapter considers a simple modification of Laffont and Tirole's (1991) standard mechanism in hierarchical centralized structures, where two agents (a firm and an agency that acts as supervisor) can collude at the expense of the principal. The firm can choose between two competing contracts: a fast contract, which is free from supervision, or a grand contract, that is subject to it. This mechanism eliminates the agency costs of supervision by providing firms with the possibility of avoiding supervision altogether in the first stage. Thus, the model yields results that are superior to the standard hierarchical model. When firms are risk averse, this mechanism also provides an insurance coverage to productive agents. As a consequence, it would be worthwhile even abstracting from collusion. Finally, the fourth chapter studies the optimal design of an organization within a principal-supervisor-agent setting, with an agent informed on the cost of his economic activity and a supervisor better informed than the principal about agent's type. This work shows that collusion is not harmful if the principal contracts with both parties and collusion is allowed only after the acceptance of the principal's contract. Moreover, supervision is valuable regardless of the intensity of asymmetric information inside the coalition. These results are robust to alternative information structures and hold for a quite generic specification of agent's type.