Triumph of the fraud(ulent) log

Fion, the representative fraudster has a natural log utility function and, depending on their individual criminal inclination, varying scalars for the “time is money” relationship. While a natural log utility function is a common, simple model used for exposition in economic theory, this financial fraud has significant negative economic effects. Consequently, governments seek to deter fraud through jail time. To ease the decision of judges in determination of the appropriate time to serve for frauds of differing amounts, the Federal government has laid out a somewhat complex mechanism – the sentencing guidelines. The development of these guidelines mandated utilization of empirical evidence as well as societal norms, a complex problem to be sure. The eventual guidelines provided no explicit utility function for a typical fraudster, rather it provided tables that allowed a judge to move from fraud amount and criminal history of the defendant to suggested time in prison. This paper uses the fraud sentencing guidelines, as promulgated, to reverse engineer the utility function as found by the Federal Sentencing Commission. It appears that, in the opinion of the Commission, the representative fraudster has a natural log utility function and, depending on their individual criminal inclination, varying scalars for the “time is money” relationship. While a natural log utility function is a common, simple model used for exposition in economic theory, this appears to be the first time it has been used in a practical and significant manner. The implications are quite clear: the sentencing guidelines presume a risk averse behavior for fraudsters, which runs counter to much empirical and theoretical work. The guidelines also encourage a “Go Big or Go Home” approach to fraud. For frauds above $7 million the suggested sentence is progressively lower than would be associated with the natural log. This is in one with Lott’s concern for the opportunity cost of wealthy individuals in prison.