Exonerations by Year of Conviction

Year of Conviction

You see it above, for 1,596 exonerations. It’s a bit more complicated than the longer-standing graph of exonerations by year of exoneration, below.

Year of Exoneration

The new graph shows when the exonerees were falsely convicted, regardless of when they were ultimately exonerated. Several convictions date back to the 1950s and ‘60s—and many are from the 1970s and early ‘80s—even though all exonerations in the Registry occurred since 1989.

The purple lines on these two graphs track the total number of exonerations for each year—of conviction or of exoneration. They tell different stories. The number of exonerations that occur per year has increased steadily since 1989 as more attention and resources have been devoted to the problem of false convictions. The underlying wrongful convictions, on the other hand, are concentrated between 1984 and 1999, and their number decreases sharply after 2007. (And then there’s a spike in 2014—we’ll return to that.)

Why are there so few exonerations based on convictions from 2007 through 2014 compared to exonerations based on convictions from 1990 to 1997? One answer is pretty clear: most false convictions aren’t exposed for many years.

That’s easy to see for homicide, the most common crime in the Registry, at 45% of the total. The time line for convictions that ultimately lead to homicide exonerations (orange) largely tracks the line for all exonerations (purple). The number per year increased sharply in the mid-1980s, stayed high for about 15 years, and then fell for convictions from the late 1990s on.

This is no surprise. The time gap from conviction for homicide to exoneration is 13 years on average, and much longer—up to 39 years—in many cases. Judging from past cases, most exonerations for false homicide convictions since 2000 will take place in years to come. There is no reason to believe that when the dust settles there will be a decrease in the rate of exonerations among recent homicide convictions.

The yearly total of sexual assault convictions that led to exoneration (red line) also increased sharply in 1980. Unlike the homicide line, however, that increase stalled in 1983 and then dropped abruptly in 1991. The average time from conviction to exoneration is similar for rape and homicide, about 13 years, but the drop in the number of convictions that produced exonerations took place nearly a decade earlier for rape than for homicide. Why, 20 years later, are there still so few rape exonerations based on convictions from 1994 and 1995?

For rape, unlike murder, it looks like that there has been a genuine reduction in the underlying rate of false convictions.

Most rape exonerations are based on post-conviction DNA evidence that proves the innocence of defendants who did not have the benefit of pre-trial DNA testing. After 1990, pre-trial DNA testing became increasingly common in America; by 2000 it was the rule across the country. Judging from the data,  pre-trial DNA testing has reduced the rates of false convictions and exonerations for rape: Rape cases are 20% of exonerations for convictions before 2000, but only 9% of exonerations for convictions since 2000; and rape exonerations that have occurred since 2000 are overwhelmingly for convictions from before 2000.

Finally, the green lines—“Other Crimes”—include everything from attempted murder to traffic violations. Most of these exonerations took place comparatively soon after conviction, on average about 5 years. As a result, the two green lines—by year of conviction and by year of exoneration—look generally similar from 1989 on, rising over time and then jumping to a record number in 2014. That spike in 2014 is an extreme manifestation of the comparative speed of these exonerations: 33 of the 36 “Other Crime” exonerees who were convicted in 2014 pled guilty to drug possession in Texas, 29 of them in Harris County, and were exonerated within a year when lab tests found no illegal drugs.