"If you look for them, you will find them."
Keith Findley, Director of the Wisconsin Innocence Project
Assistant Professor of Law at the University of Wisconsin Law School

There was a time when we just thought only guilty people were convicted. To think otherwise was, in some corners, to literally engage in fantasy.

Over recents years, however, facts and the laws of probability have updated our perception and knowledge. The science of DNA testing has indisputably proven scores of citizens actually innocent. And we have come to embrace a broader understanding that humans (and the systems they design) are not, and can never be, perfect.

The chart below illustrates the potential scope of the wrongful conviction problem in Texas. We apply various error rates (ER) to the total number of felony convictions in Texas (most recent period is September 14, 2014 to August 31, 2015) to get a range of potential wrongful convictions annually. We plot that number against the known number of Texas exonerations (average from 2011-2015). Additional details about how we constructed this chart can be found below the chart.

There are two unavoidable takeaways from this illustration. First, due to the volume of felony convictions in Texas (108,405 in the most recent year) even a small error rate has enormous impact on the lives of affected citizens and their families. Second, the gap between wrongful convictions being generated and the actual number of wrongful convictions being corrected through exonerations is significant, even at lower error rates. Indeed, the gap is so large you can hardly see the bar representing the five-year Texas exoneration average (22).

The Potential Scope of Felony Wrongful Convictions in Texas

Sources: 2015 Annual Statistical Report for the Texas Judiciary; National Registry of Exonerations

First, a note about wrongful conviction error rates (the numerator in the calculation). The actual error rate for wrongful convictions can never be known. However, over the years, there have been several studies done that point to an error rate range of one half of one percent to over ten percent. For our chart, we use the lower range of the error rate estimates that have been documented. Also, we apply the error rate only to felony convictions, which have been the focus of wrongful conviction error rate studies.

Second, a note about the population to which we apply the error rates. We use only felony convictions in Texas as documented by the Texas Office of Court Administration (OCA). Those felonies include homicide, assault, robbery, burglary, auto theft, drug offenses, DWI, and "other" felonies. We exclude misdemeanors, again because felony convictions have been the focus of wrongful conviction error rate studies. The "baseline" number for total felony convictions in Texas used for this chart is 108,405. So for example, 108,405 x 1.0% error rate = 1,084 potential wrongful convictions.

To summarize, we know the number of felony convictions each year. We know the number of actual exonerations through the National Registry of Exonerations (if anything, this number is likely low due to underreporting). We don't know the exact error rate, but we know there is one. More importantly, we now know that even a small error rate tells us that there is much cause for concern and much more work to be done.