In 2001, murder convictions accounted for about 13% of the prison population, but 85% of exonerations (166/196).
In 2001, death row accounted for about 1/4 of 1% of the prison population, but 22% of exonerations (74/196).
What accounts for this abnormally high % of exonerations from these two groups? Only two possible explanations exist:
1. Wrongful convictions are not more likely to occur for murder convictions, especially death-row cases, but are more likely to be discovered because of the higher attention paid to these cases in post-conviction review. If that is true, if we paid the same attention to other convictions, we could expect that "there would have been over 29,000 non-death row exonerations in the past fifteen years rather than the 265 that have in fact occurred."
2. Wrongful convictions are more likely to occur for murder convictions, especially death-row cases. This may result from several causes:
- the extraordinary pressure to secure convictions for heinous crimes
- the difficulty of investigating homicides because, obviously, the victim cannot provide information
- extreme incentives for the real killers to finger or frame an innocent person
The study concludes that it is likely a combination of both that accounts for an abnormally high percentage of exonerations coming from murder convictions, and especially death-row cases.