False Confessions Predict a Delay Between Release From Incarceration and Official Exoneration

Objectives: Little empirical research has examined postconviction processes associated with the unique legal events of release from incarceration and official exoneration. Across various models, we tested the influence of risk factors associated with wrongful convictions (false confessions, faulty or misleading forensic evidence, inadequate legal defense, mistaken eyewitness identifications, official misconduct, and perjury) and relevant alternative factors (e.g., presence of DNA, false guilty pleas, and race) on the exoneration process, with a particular focus on the role of false confessions. Hypotheses: We expected that all risk factors would be meaningfully associated with the duration between wrongful conviction and release but that false confessions would be associated with longer delays between release and exoneration and would remain a meaningful predictor of the delay even when accounting for alternative factors. Method: Using data from documented exonerations of murder, attempted murder, and accessory to murder in the National Registry of Exonerations (N = 1,074), we examined the association of risk factors and alternative predictors with the time between exonerees’ wrongful conviction and release from incarceration and the time between release from incarceration and official exoneration. Results: Overall, five of the six risk factors predicted the time between wrongful conviction and release from incarceration, but of the risk factors, only false confessions predicted the time between release and exoneration (d = 0.28; 95% CI [0.13, 0.43]), even when we controlled for relevant alternative factors (d = 0.29; 95% CI [0.14, 0.43]). Conclusions: Exonerations that involve false confessions are associated with delays in the critical window between innocent people’s release and official exoneration—a time during which these innocent people are precluded from accessing reintegration aids and may struggle to find housing and employment.