That is because any effort to estimate it involves extrapolation from just two numbers, neither one satisfactory. There have been 214 exonerations based on DNA evidence, almost all of them in rape cases, according to the Innocence Project at the Cardozo School of Law. But there is no obvious control group to measure these exonerations against.
Virginia, though, has discovered thousands of closed rape files from 1973 through 1988, many with untested biological evidence. DNA testing of a preliminary sample of 31 of them yielded two wrongful convictions. Those numbers are too small to be reliable, of course, but they would suggest a false conviction rate of 6 percent.
Even that rate may be low, said Shawn Armbrust, the executive director of the Mid-Atlantic Innocence Project. Ms. Armbrust said investigators in Virginia were able to get results in only 22 of the 31 tests, suggesting a false conviction rate of 9 percent.
The other important number comes from death row. According to the Death Penalty Information Center, 127 death row inmates have been exonerated.
Here we do have a control group. There have been more than 7,000 death sentences since the Supreme Court reinstated the death penalty in 1976.
But exoneration in the capital context is a funny concept. It suggests complete vindication, but its real meaning is generally narrower.
DNA evidence in a rape case can provide something like categorical proof of innocence. Death row exonerations, on the other hand, can be based on all sorts of things, like, say, prosecutorial misconduct. In other words, it is possible to wrongfully convict a guilty defendant.
Mr. Marquis, the Oregon prosecutor cited by Justice Scalia, says the number of authentic death row exonerations is more like 30. Many people exonerated in the legal sense, he said, in fact committed the crime but could not be proved guilty beyond a reasonable doubt.
Professor Gross thinks the number of guilty people released from death row is very small.
Professor Gross concluded that the false conviction rate for death row inmates has ranged from 2.3 percent to 5 percent. Were even the lower end of that range applied to people who received prison sentences of a year or more in the last three decades, he wrote, it would suggest that about 185,000 innocent people have served hard time.
But extrapolating from capital crimes to felonies generally is problematic whatever the number of exonerations.
Consensus on Counting the Innocent: We Can’t [NYT]