Seminar by Brishti Guha (Jawaharlal Nehru University)

This is the first paper to study accomplice plea bargains when motivated jurors incur an effort cost.  I find that a prosecutor who wishes to ensure punishment for the guilty, and acquittal for the innocent, induces a unique perfect Bayesian equilibrium which is fully separating, simultaneously minimizing court costs as well as jurors’ cognitive efforts. In this equilibrium, a very harsh plea bargain (with negligible plea discounting) is offered, and is accepted by all guilty defendants, and rejected by all innocent ones. Though the prosecutor has the ability to induce semi-separating equilibria, he will prefer the fully separating one. We can use the Cho-Kreps intuitive criterion to show that a pooling equilibrium does not exist. My results contrast sharply with the plea bargaining literature which finds a tradeoff between sorting efficiency and the harshness of punishments meted out to the guilty and which does not support a fully separating equilibrium.