This article is the first systematic empirical study of how the American criminal justice system discovers and responds to factual error based on actual innocence. The study analyzes a data set of 260 cases of wrongful conviction of the innocent and 200 near misses (i.e., dismissals and acquittals involving an innocent defendant) to better understand the sources of and bases for exoneration; who is responsible for, as well as who opposes, exoneration; the statistical correlates of exoneration; and the primary methods and mechanisms involved in the path to exoneration.

This study leads to several findings. First, wrongful convictions are difficult to reverse in the absence of dispositive evidence of innocence. The vast majority of exonerations relied on one or two bases, and even then most required DNA evidence. Second, the adversarial nature of the criminal justice system continues from the trial level to subsequent efforts to exonerate the innocent. Police and prosecutors maintain their roles, infrequently playing a central part in investigating or advocating for exoneration and serving as the largest combined source of opposition to exonerations. Finally, exonerations take a long time, even longer when based on DNA evidence, which appears to be the primary basis for clearing defendants.

After examining these findings, the authors advocate for the following changes in the United States criminal justice system: 1) police and prosecutors must take a more active role in the review and reversal of factually erroneous convictions; 2) additional juridical proceedings are needed for the wrongly convicted to prove their innocence even after conviction; 3) efforts must be made to prevent wrongful convictions at the front end because the resources for freeing the wrongly convicted are so limited and the path to exoneration following conviction is filled with formidable challenges.