The Unintended Consequences of Raising the Standard of Proof in Criminal Sentencing

• In Event: Economics of Crime: Arrest & Sentencing

Abstract

We study how raising the standard of proof affects charging and sentencing outcomes in criminal courts. Procedural rules are often designed to constrain discretion and improve fairness, yet their effects depend critically on how legal actors respond. We exploit a Supreme Court decision that increased the evidentiary burden required to impose sentencing enhancements, allowing us to examine how prosecutors and judges adjust behavior when one margin of discretion is restricted.
We begin by documenting a sharp decline in the use of sentencing enhancements following the increase in the standard of proof. Because enhanced sentences are mechanically longer than non-enhanced sentences, this change should reduce average sentence lengths. However, we find the opposite pattern: average sentences increase relative to the pre-reform period. This divergence from the mechanical prediction suggests that legal actors adjust behavior along other margins.
To identify the source of this response, we examine prosecutorial and judicial behavior separately. Using rich administrative data and a regression discontinuity design, we estimate the effects of the reform on 13 outcomes related to prosecutorial charging and plea bargaining. Across all measures, we find precise null effects, indicating no systematic prosecutorial response. In contrast, we uncover substantial heterogeneity in judicial behavior. Judges with greater baseline leniency significantly increase sentence lengths following the reform, while stricter judges reduce sentence lengths, leading to a compression of sentencing outcomes across judges.
We show that this pattern is consistent with institutional constraints embedded in sentencing guidelines. Stricter judges are more likely to be operating near the upper bounds of guideline ranges and therefore have limited ability to offset reduced access to enhancements. More lenient judges, by contrast, retain discretion to increase sentences within the guideline window and appear to do so in response to the reform. As a result, discretion is not eliminated but reallocated across margins, generating both an increase in average sentence lengths and a reduction in dispersion.
We interpret these findings through two complementary mechanisms. First, reduced use of enhancements may limit the information available to judges about case severity, encouraging more pooled sentencing decisions. Second, and more strongly supported by the data, judges may respond to the loss of discretion over enhancements by increasing punishment along remaining discretionary margins.
Our results highlight that tightening procedural constraints can generate unintended and countervailing effects, underscoring the importance of accounting for behavioral responses when evaluating institutional reforms.

Author

• Spencer Cooper, University of Connecticut
  Presenting Author