Trends and Data in Exonerations: An Interview with Samuel Gross of the National Registry of Exonerations

Infographic 2106 Exonerations

No one wants to think about the possibility of an innocent person being wrongfully convicted.

Blackstone Image Quote
Image Credit: Wikimedia Commons

From Blackstone to Benjamin Franklin; Voltaire, to even the book of Genesis, it’s a long-held belief that it’s better to “err on the side of innocence” by allowing a number of guilty people to go free, than to wrongly convict one innocent defendant.

But, still, it does happen.

And when it does, the National Registry of Exonerations is there to not only count those instances, but to also help tell their stories – and to learn from them what we can.

We sat down with Professor Samuel Gross, of the University of Michigan Law School, editor of the National Registry of Exonerations, a joint project of the Newkirk Center for Science and Society at the University of California Irvine, the University of Michigan Law School and the Michigan State University College of Law, to learn more about the Registry, the data they collect, how the data can be used, and what trends he sees.

(This interview has been edited for length and clarity.)


Justice Clearinghouse Editors (JCH): So let’s start off with just a little bit of the history of the National Registry of Exoneration. How it started, and how it developed?
Dr. Samuel Gross
Photo Credit: Dr. Samuel Gross, University of Michigan, National Registry of Exonerations

Professor Samuel Gross: In 2005, I, together with several students, published an article in the Journal of Criminal Law and Criminology that included what was at the time the largest database of exonerations that was available – around 240 cases. But by the time it was published, I knew it was incomplete because I kept hearing about the exonerations that weren’t included. When we wrote the article, we said we’d update it, but it became clear that updating it didn’t make any sense because there was nothing like a complete list of exonerations that had occurred, not close to it.

A few years later there was a possibility that other researchers, friends of mine, would put together something that they called an Encyclopedia of Wrongful Convictions, basically a hard-copy encyclopedia that collected all known exonerations in America. But that didn’t happen. By 2007 nobody was producing new printed encyclopedias of any sort,  and in any case, it was a big task.

By 2009, I realized if anything like this was going to happen, I would have to organize it myself. Which I did, over a period of time, together with Rob Warden, then the executive director of the Center on Wrongful Convictions at the Northwestern University School of Law.

We put together a small amount of money, and, working with law students and others, got the registry together and released it publicly on our website in May 2012.

In May 2012, we listed 891 individual exonerations. We now have 2,018 [April, 2017]. We have a staff of 2 and we’re about to add another person, and the number of editors have grown from just me to me and three others.

For a while, the Registry was a joint project of the University of Michigan Law School and Center on Wrongful Convictions at Northwestern University School of Law.  Now, as I said, it’s a joint project centered at the Newkirk Center for Science and Society at the University of California Irvine, together with the University of Michigan Law School and the Michigan State University College of Law


JCH: How do people typically use the registry?

Samuel: I like saying that the Registry does two things: we tell stories and we count things.  And they’re equally important.

For each of the individual exoneration that we list, we have a narrative summary – the story of the case – and if possible, a picture of the exonerated person, and basic information about them.  We also have a large amount of data that is coded for each case. Many of those items are displayed on the website in a format that permits searching, filtering and sorting.


JCH: You started back in 2005 and you’ve been tracking these cases ever since. Last year was a record year for exonerations. How did 2016 fair in comparison to 2015?
Year and Type of Crime Exonerations
Graphic Credit: National Registry of Exonerations

Samuel: If you look at our graph of Exonerations by Year and Type of Crime you see that the number of exonerations has been going up steadily for years. We’ve had record numbers of exonerations for three years in a row: 20142015, and 2016.

Each year we add cases, some of which occurred in the year in which we added them, and some of which occurred years before. Over the past several years we’ve gotten more efficient at finding exonerations, so we learn new exonerations closer to the time when they happen. But we’re always adding cases that occurred at any time since 1989.


JCH: What are your observations about the differences between 2015 and 2016?

Samuel: There are a couple of noticeable differences between the two years: 2015 had a record number of homicide exonerations, 2016 did not. That’s hardly surprising.

As you see on the graph for Exonerations by Year and Type of Crime, the orange line goes down a little from 2015 to 2016. On the other hand, “Other” crimes – mostly drug possession cases – continues to go up, and Child Sex Abuse went up a good deal for a relatively uncommon exoneration category, from 7 in 2015 to 17 in 2016.

I think those changes are just fluctuations: they don’t mean anything in particular. This is a reasonably continuous process with bumps up and down.

The exonerations in both 2015 and 2016 had high rates of cases with official misconduct. I think that may represent a trend, but we’ll see if it is over time.

Remember, the people who were exonerated in 2016 were not convicted in 2016. The convictions occurred on an average about 8 and a half to 9 years earlier. But the average doesn’t really tell you that much, because those the convictions occurred across a wide range of years. Someone who was exonerated in 2016 might have been convicted in 1970 or in 1985 or in 2015.

We’re not picking up any change in the frequency of the official misconduct in convictions that occurred that far back. I think what this means is that nowadays we’re more likely to discover official misconduct in cases that occurred years ago.  At the same time, courts and the prosecutors now seem more willing to acknowledge it than they were in the past.

If you look at the two reports we issued in 2015 and 2016, you’ll see that each of them shows a substantial increase in the number of exonerations that involved work by conviction integrity units in prosecutors’ offices. That’s a genuine and important recent trend. More Conviction Integrity Units are being created each year, and more exonerations are attributed to their work.  This could be the beginning of a big change.

Infographic 2106 Exonerations
Click here to see full Infographic


JCH: And when you say “official misconduct,” can you give me examples of what you’re talking about?

Samuel: Official misconduct is generally by prosecutors or police officers, in about equal proportions. Less frequently, it is misconduct by other government workers. The next most common category are forensic scientists and child welfare workers, but in the great majority of cases, it’s either prosecutors or police officers or both.

The single most common type of official misconduct that we run into is withholding exculpatory evidence, which is usually done by prosecutors – not informing the defense or the court of evidence that would help the defense and support the defendant’s claim of innocence. This happens in about 35% of all exonerations.

The next biggest category of misconduct is a collection of things that we group together as “witness tampering.” This can include coercing witnesses to say things, or bribing them, or manipulating them into saying things or in extreme cases suborning perjury. That type of misconduct is overwhelmingly done by police officers.

Then there are a bunch of other things that count as misconduct, but none of them is as frequent as those two.


JCH: Why is it that we’re seeing more exonerations?

Samuel: My view is that it’s a combination of two things. First, more resources are devoted to finding them and advocating for the defendants who may be innocent. More innocence organizations are doing it, and, in particular, more conviction integrity units in prosecutorial offices.


“[N]owadays we’re more likely to discover official misconduct in cases that occurred years ago.  At the same time, courts and the prosecutors now seem more willing to acknowledge it than they were in the past.” 


Samuel Gross, editor
National Registry of Exonerations



Conviction Integrity Units, especially the ones in Harris County, Texas, and Brooklyn, (King’s County), New York, DAs’ offices, have been responsible for many of the exonerations in the last few years.

That means there are more people working to get innocent defendants released as exonerations have become more common, and information about them has become more widespread. As a result, courts, prosecutors, all of us – including defense attorneys, police officers, probation officers, the general public – now realize that it’s not that surprising to find out that someone who has been convicted of a crime may actually be innocent. 

JCH: So, what is the role of technology in exonerations? Does that play a role in the increasing number of exonerations as well?

Samuel: It has had a huge role.

If you look at some of the graphs I was pointing to, one of the graphs there shows DNA exonerations in separate lines. What you see may surprise most Americans. If you look at DNA exonerations by year, what you’ll see that DNA exonerations have always been a minority of exonerations. Recently, they have been a decreasing minority because the number of DNA exonerations has remained relatively flat since about 2000, more or less, while the number of non-DNA exonerations has increased sharply.

That’s not surprising. DNA is extremely valuable in a case in which the criminal left biological trace evidence that identifies the person who committed the crime. That happens most often in rape cases, where semen is left, and also in some other violent crimes, especially murders, when blood left is behind that identifies a particular person who did the crime.

In most crimes, however – drug crimes, robberies, burglaries, assaults, not to mention fraud or other white collar crimes – DNA has no useful role, either in proving that the defendant is guilty or in exonerating a defendant who is innocent.

When DNA exonerations began in 1989 -which is the first year we included in the registry – they sparked a revolution in how we think about false convictions. They showed that quite a few defendants who had been convicted of sexual assaults and some murder cases – in trials that nobody ever questioned, where the evidence against them seemed perfectly clear – turned out to be innocent. What’s more, the proof of their innocence didn’t depend on believing witnesses you might dispute, but was the product of a scientific technique that was highly accurate, widely trusted, and was not known in the time of the original conviction so it was not possible to get that information in 1975 or 1982 or whenever they went to trial.

Over time, that deeply changed the way we view the whole issue. Some people described DNA as a sort of as a flashlight, that suddenly became available and made it possible to see things in a dark basement: all of a sudden we found all sorts of broken junk we didn’t know about. That, more than anything else, led to new attention to the problem of false convictions, increased our interest in the problem, and the changed our view of the danger of convicting innocent people.

At the same time this was going on, there was also a large increase in the number of exonerations of defendants who had been sentenced to death, starting in the mid-1970s. Most of those exonerations did not involve DNA evidence. That started before 1989.

So in the 1990s you had a large number of cases which DNA had proven the innocence of people who basically had convictions that nobody really questioned – exonerations that seemed to come out of the blue. And in parallel and overlapping, there were a bunch of cases of defendants who were sentenced to death and years later, on careful review, turned out not to be innocent.



“We’re looking for [exonerations] more carefully

and recognizing them more willingly.”


Samuel Gross, editor
National Registry of Exonerations


Those two trends together, I think, are responsible for many of the changes we see now. Of the two, I think the advent of DNA evidence is the more important one. I believe it is indirectly responsible for many of the non-DNA exonerations that have followed,  both by increasing attention and resources devoted to the issue, and by making everybody realize that this is a real problem, so we have to take seriously the possibility that we’ve convicted people in error.


JCH: How could the Justice Clearinghouse members use your database? What are the things they can learn or glean from your database that might be useful to them?

Samuel: Well, if they’re interested in how false convictions occur, then there’s many things they can learn. We have the most complete data that exist and they’re easy to retrieve.

Interactive Data National Registry of Exonerations
Graphic Image: Interactive Data National Registry of Exonerations

For example, if you go to Interactive Data Display, you can find all sorts of things.

So if you just look at all these cases, and choose a category, say Drugs: it has 236  exonerations [in March 2017], with an average of 1.3 years lost to imprisonment per exonerated defendant.  If you look down below, you see how many occurred by year, and you’ll see that these drugs cases are mostly recent – and mostly from Harris County, Texas, where the Conviction Integrity Unit cleared a lot of false drug convictions since 2014. And then if you hover over those bars, the orange bars, you notice that each layer is a case and it shows you which one it is.

There are a lot of things like these that you can easily find if you’re interested.

There’s a basic lesson, which I mentioned before: The number of exonerations is increasing. What that represents is an increasing awareness of the scope of the problem. It doesn’t mean that more people have been wrongly convicted in the last few years; these exonerations are for wrongful convictions that are years or decades old. It means that we’re looking for them more carefully and recognizing them more willingly. That’s the most important lesson.

This is a serious problem. Maybe 2% of convicted criminal defendants are innocent, maybe it’s 4% or 3% or criminal convictions – we don’t know – but it’s a lot of people. Even if it’s 1%, that’s tens of thousands of innocent defendants convicted each year across the country. The exonerations that we know about are only a small proportion of the wrongful convictions that occur.