Criminal network, leniency, and market externalities

We analyze the self-reporting incentives fostered by a leniency program within a criminal network formed by a supplier of an illegal good and his dealers who compete against each other in the product market. We show that when it is viable, a first-informant rule always performs better than an all-informant rule—that is, it induces a lower level of crime. Nevertheless, the viability of a first-informant rule may be compromised if the baseline probability of conviction is sufficiently low, thereby placing disproportionate reliance on leniency over other investigative efforts for securing convictions.