The one modern case that death penalty opponents point to that comes closest to meeting this claim involves Claude Jones Texas, who was convicted in 1989 and executed in 2000. Two pieces of evidence were involved in his trial: the testimony of two accomplice who planned the robbery with Jones and who provided him with the gun used in the killing as well as a hair sample at the scene that turned out to be someone else’s. The fact that the hair at the scene was someone else’s would have meant that piece of evidence could no longer be used to prove that Jones actually went into the store, but the failure of the DNA test to match Jones to the hair does not prove that Jones didn’t do the killing because some many people visited the store it could have been from someone who didn’t commit the murder. The issue here is whether there was still enough evidence with the witness’ testimony and the gun to place Jones at the scene of the crime.
Some discussion on the error rate in death penalty cases is in order. Antonin Scalia noted this in his concurrence in KANSAS v. MARSH (No. 04-1170) 278 Kan. 520, 102 P. 3d 445:
“[L]et’s give the professor the benefit of the doubt: let’s assume that he understated the number of innocents by roughly a factor of 10, that instead of 340 there were 4,000 people in prison who weren’t involved in the crime in any way. During that same 15 years, there were more than 15 million felony convictions across the country. That would make the error rate .027 percent—or, to put it another way, a success rate of 99.973 percent.” The Innocent and the Shammed, N. Y. Times, Jan. 26, 2006, p. A23. . . .
There are some issues that need to be taken into account in Scalia statement. Surely, assuming that the number of innocents is over 10 times great than the 340 noted offsets some errors in the analysis. However, there are some significant issues that must be addressed.
As of May 13th, the Innocence Project puts the number of those exonerated by DNA evidence at 316. The one problem with the New York Times discussion is that the correct denominator really isn’t 15 million since DNA evidence isn’t available in all those cases. For murder cases, a review of 400 murder cases found that 4.5 percent had DNA evidence. However, that is actually an extremely conservative figure because many of the 95.5 percent of the cases never go to trial. Between 2000 and 2008, 62 percent of murders were cleared through arrest. Suppose that there is a 90 percent conviction rate conditional on murder (during the late 1980s and 1990s the conviction rate was as high as 93 percent). Obviously cases with evidence were much more likely to end in arrest or conviction.
So assume that this sample of 400 murder cases was random, about 223 would have resulted in convictions and 18 of those would have had DNA evidence. That raises the rate of DNA evidence to 8.1 percent. The true rate of DNA evidence is thus someplace between 4.5 and 8.1 percent.
If there were 260,000 Americans convicted of murder between 1989 and 2011, there was something between 11,700 and 21,060 cases with DNA evidence. Assuming that in 53 cases convictions over that period were overturned (see both Tables shown below with data from the Innocence Project), that implies an error rate of 0.25 to 0.45 percent.
Even this rate significantly overestimates of the true error rate because many of these convictions involved murders committed well before 1989. If one limits the cases to murders committed after January 1989, there are 38 cases and an implied error rate of 0.18 to 0.32 percent.
The number of people charged with the death penalty for murder who were later cleared because of DNA evidence is 34. If one uses that number, the error rate of 0.16 to 0.29 percent. Obviously from the preceding discussion there are biases going both ways, though mainly ones that overestimate the relevant number in the post January 1989 period. (UPDATE: Radley Balko at the Washington Post questioned the source of this figure. On that point, it is right at the top of the page that I linked to, though it does require addition: “18 of the 316 people exonerated through DNA served time on death row. Another 16 were charged with capital crimes but not sentenced to death.” 34 = 18 + 16.)