'Texas criminal justice system has major flaws': Estimating innocent people in prison

The title of this post is the headline to a Corpus Christi Caller Times editorial reacting to the slew of recent DNA exonerations in Texas as well as the Cameron Willingham case. The sentiments expressed, it seem to me, represent a growing consensus that Texas justice continues to suffer from serious problems.

Despite the alarmist headline, though, it struck me that some readers of the editorial and other news coverage of innocence cases may be inclined to minimize the problem evidenced by 38 DNA exonerations and one falsely convicted man who was executed.

To understand the scope of the actual innocence problem one must put these numbers in context. I've written frequently about efforts to estimate statistical rates of actual innocence in Texas, but let's try and get at the numbers another way.

According to the Caller Times, there have been 38 Texas DNA exonerations. (You hear counts of 40 or higher including more recent exonerees, but the Innocence Project doesn't include them as exonerated until the Court of Criminal Appeals finally clears their name.)

But 20 of those 38 have come out of Dallas, with just 18 in the rest of the state. That's not because Dallas had a worse problem than other jurisdictions, necessarily, but because Dallas kept old rape kits from past decades and virtually nobody else did.

What's more, DNA evidence only exists in about 10% of violent crimes in the first place, so in many cases where innocent people have been falsely convicted, there's simply no exonerating forensic evidence available.

So to reach a back-of-the-envelope estimate, here's the current breakdown of innocence cases:

• Dallas: 20
• Rest of State: 18

Since DNA evidence is only available in 10% of violent crimes, let's assume that innocent people among those 90% cannot be exonerated by DNA and remain in prison. That would bring the total of convicted innocents exonerated and in prison to:

• Dallas: 20o
• Rest of State: 18o

But Dallas kept their old evidence, while the rest of the state mostly threw it out in older cases. It's probably generous to assume most jurisdictions kept biological evidence in even 10% of old cases, even more so to believe that all the old evidence has been tested. But even if both those unlikely assumptions are true, that would still mean our estimate for the likely number of innocent people in the rest of the state goes up by another factor of ten, or:

• Dallas: 200
• Rest of State: 1,800
• Total: 2,000 innocent people in prison or recently exonerated
  

Notably, that gets us right into the ballpark with other possible methodologies of estimating innocence.

Josh Marquis of the national District and County Attorneys Association used his own back-of-the-napkin methodology to estimate an actual innocence rate of .75%, which would be about 1,200 prisoners in Texas. Looking at the rate of Texas death row exonerations and applying it to the whole prison population would get you around 2,400 innocent prisoners. Applying other innocence rates estimated from various sources to Texas' large prison population gets estimates as high as 3,500 to 5,000 innocent prisoners out of those currently incarcerated.

All of these estimates, I've argued, may understate the false conviction rate by excluding drug war cases. And whatever rate one decides is fair to apply to the prison population, the rate among probationers is likely a little higher because of the incentive innocent people have to take a deal to avoid incarceration.

These false convictions are happening for specific, often repetitive reasons: The most attention has been focused on the failure by police to use best practices for eyewitness identification, but there are a litany of other contributing factors (which regular readers could likely recite) like mendacious informants, goal-oriented forensics, false confessions, and occasionally even police and prosecutor misconduct. On some of these topics, the Texas Lege has taken baby steps. On most of them they've done nothing.

So yes, 38 DNA exonerations in Texas (or whatever is the real number) are a serious concern. But what's outright alarming is that these men represent hundreds or even thousands of other innocent people sitting in Texas prisons who modern technology cannot liberate.

UPDATE: I've added a reader poll in the sidebar listing the various actual innocence estimates I'm aware of as applied to Texas' prison numbers - register your vote how many innocent people you suspect are currently residing in Texas prisons. If you calculate things differently, pick the answer that's closest and let me know how you think the numbers should be crunched in the comments of this post. CLARIFICATION: In the reader poll, treat "5,000 as "5,000 or more." I'd change the text but the software won't allow it once voting has begun.