On criminal justice sentencing, structural racism and policing reforms with leading scholar Brandon Garrett

Emilia Chiscop-Head, PhD - Scientific Integrity Associate, ASIST

Brandon L. Garrett, JD is the inaugural L. Neil Williams, Jr. Professor of Law and came to Duke in 2017 from the University of Virginia, where he was the White Burkett Miller Professor of Law and Public Affairs and Justice Thurgood Marshall Distinguished Professor of Law. Garrett is a leading scholar of criminal procedure, evidence, and constitutional law. His research areas include forensic science, eyewitness identification, corporate crime, civil rights, habeas corpus, and criminal justice policy.

Garrett’s work has been cited by the US Supreme Court and other courts in the U.S. and internationally. He has initiated law reforms, including the American Law Institute’s project on policing, for which he serves as Associate Reporter. Garrett maintains online data sets that are made available to the public and are being used in criminal justice practice and education, including: “End of Its Rope: Data on Death Sentencing,” “Corporate Prosecution Registry” and “Convicting the Innocent: DNA Exonerations Database.

The interview below focuses on his 5th book, “End of Its Rope: How Killing the Death Penalty Can Revive Criminal Justice” (Harvard University Press, 2017) and the “Autopsy of a Crime Lab,” which will be published in spring 2021. Garrett reveals evidence-based, insightful and timely details about systemic racism in the criminal justice system and policing reforms that he suggested.


What inspired you to write “End of Its Rope”?

Brandon Garrett: In 2017, I published “End of Its Rope,” a book that explores what can explain the remarkable drop in death sentencing in the United States. As a law professor, I knew that the rules for how you impose death sentences had not changed. When I was a law student first learning about criminal law, death sentencing had reached its modern height – over 300 sentences a year. In states like North Carolina, Virginia, Texas, it was routine to have dozens of death sentences in a single year. I did not anticipate that death sentencing in this country was about to fundamentally change, and enter into a freefall, which has continued for two decades now. In my book, I explored why this happened. A lot of data collection needed to be done because there were not case-level death sentencing information from the 1990s to present. Over several years, with a team of law students who worked incredibly hard, we were able to build a data set that included death sentencing from 1990 through the present. We've made the data available on a research website.

The goal of the book was to use that data as a starting point to unpack what are the different factors that can explain this remarkable modern death penalty decline, and what are the lessons can we learn from the fall of modern death sentencing.

Has your research confirmed the hypotheses?

Brandon Garrett: Some of the findings confirmed initial hypotheses and others were a surprise. For example, in prior work, I had studied wrongful convictions. In fact, at the Duke Wilson Center for Science and Justice, we're updating and maintaining a website with data concerning DNA exonerations. There have been over 370 DNA exonerations in the country, while in North Carolina, we have seen a number of high-profile exonerations. I wondered whether DNA exonerations in death penalty cases have changed death sentencing practices. I thought that people could have turned away from the death penalty because of the fear that an innocent person could be executed. Instead, I found no correlation between the number of exonerations and state-level death sentencing decline. In fact, it was the opposite. Where you have death sentences there are exonerations, and the states with most death sentences, like Florida, also have the most exonerations. The same factors that produced lots of death sentences in some states may have produced more exonerations of innocent people sentenced to death. That was a sobering finding.

The strongest and most powerful effect that my colleagues and I found, had to do with the role of lawyers: States that created an office to handle death penalty trials saw much steeper declines in death sentences. I suspected that it might be an important factor because, traditionally, death penalty trials were extremely short (one or two days), with no witnesses, and lawyers who handled the most serious criminal cases had very little or no training. When states funded offices, with defense lawyers with experience doing death penalty work, the result was quite striking.

We also found a very powerful effect on the prosecution side. My colleague and coauthor Lee Kovarsky has termed it a “muscle memory” effect: counties that have prior death sentences are much, much more likely to impose death sentences in the future. We saw a strong path dependency even if homicide rates go up, even if they go down, even if the population of the county changes the income of the county.

Traditionally, many of these district attorneys made no secret of the fact they ran for office saying “I'm pro death penalty, I will seek it whenever I can.” Whereas other prosecutors made a point of saying that they were quite cautious about seeking the ultimate punishment.

The result is that the death penalty became a county-level phenomenon. We do not have “death penalty states,” so much as a scattered group of “death penalty counties.”

How have the findings of the book been used?

Brandon Garrett: Researchers, lawyers, and policymakers have used the data. I shared the data with the Death Penalty Information Center, as they were building an expanded research hub. Other researchers have used it as well. It has been and can be used in cases in which people are litigating claims concerning the patterns in death sentencing. One of our goals was also to make this broader point that although local democracy is really important - one of the reasons we elect people at the local level is to have accountability to the voters - there can be real costs when political preferences drive the most of death sentencing outcomes. If you have arbitrariness even in death sentences, it raises real questions about why we are doing more broadly in our criminal system. One of the tasks, since the book came out, has been to build out research to look at other types of serious sentencing in this country. At the Duke Wilson Center for Science and Justice we've conducted two studies so far and we have others in the works looking at non-death sentencing using some of the same techniques, examining individual level, at county level, prosecution district level disparities. We are also researching “life without parole” (LWOP) sentences which have exploded in the past years.


There was no public database on who had been sentenced to death. There was no public database describing the features and characteristics of DNA exonerations in this country. The work is interdisciplinary and it has involved undergraduate students, graduate students, postdoctoral fellows, and faculty in multiple disciplines


How do you administer the databases?

Brandon Garrett: At the Wilson Center, we always make data available to the public. Our goal is to provide a public good. We want other people to understand the findings, engage with the data, do further research, and benefit from our work. In my career, I have always been troubled that information about some of the most serious and compelling criminal cases is lacking.  There was no public database on who had been sentenced to death. There was no public database describing the features and characteristics of DNA exonerations in this country. The work is interdisciplinary and it has involved undergraduate students, graduate students, postdoctoral fellows, and faculty in multiple disciplines. We are also working with the Law Library at Duke to host some of our data sets.

We have also created web interfaces to make out data collections particularly accessible and interactive. For example, we recently created, with the help of Professor Maria Tackett of Duke Statistics and undergraduate statistics student Caroline Levenson, a Shiny app (a data visualizations app) for displaying data on criminal debt in North Carolina. This website is a remarkable resource that allows one to see the scale of paid and unpaid criminal debt, by county in North Carolina, in what types of cases. This resource has already been used by journalists in their reporting and by lawyers in litigation. We also produced to report to describe what we had found – that 1 in 12 North Carolinians have unpaid criminal debt - and described the implications of these findings.

Do you also send reports to policy makers?

Brandon Garrett: One of our goals is to do research in the public interest, and to translate our research in a way that is useful to policymakers. Some of that work involves collaborating with policymakers directly and talking to them about what information would be most useful to them. We have created short, one page fact sheets with visuals, synthesizing our findings. We have also created longer reports with more detailed policy recommendations. We have written op-eds and engaged with journalists. Accompanying data with visuals and ‘take away’ points can be effective. Our audiences want to know what are the reforms that they should consider. What are the possible ways of addressing a problem?

Another way that we can affect decision makers, is to lend our research as part of a brief in a case.  Judges have a practice of allowing outside experts that aren't part of the litigation to file a brief called an “amicus brief.” We have filed quite a few amicus briefs, often signed by and with the participation of leading scientists. The goal is often to make a research-based point regarding a question in a case.


It is not enough to adopt narrow reforms, like a ban on “chokeholds” or altering the liability of individual police officers. There are deep-rooted structural problems and you cannot have a checkbox mentality towards solutions


Could you offer one example of the most important reform or law that you know you suggested or impacted with your research?

Brandon Garrett: A group of us at the Wilson Center, collaborating with colleagues at other law schools, issued recommendations on policing reform this summer. Criminal justice problems are often complicated enough that there is not always a simple solution.  We included dozens of reforms that we thought were collectively really important. One point was that it is not enough to adopt narrow reforms, like a ban on “chokeholds” or altering the liability of individual police officers. There are deep-rooted structural problems and you cannot have a checkbox mentality towards solutions.

There are some problems for which a single simple reform can make a big difference, though. One example is driver's license suspensions in states like North Carolina. We think that the solution is simple: states should not be able to suspend driver's licenses for non-driving related reasons. There are other ways to collect unpaid debt, and states should be thinking about whether a person can afford to pay it at all. The weapon to enforce debt collection should not be depriving someone to have the ability to legally drive, which is necessary for people's lives and livelihoods.


It takes a broader social and political movement to create action around longstanding racial disparities


You discussed in your book, structural racism and other types of biases. Could you tell us more?

Brandon Garrett: There are vast racial disparities in how people are treated in the criminal system, from initial police stops, to arrests, to sentencing. I described some of the work we've done on the most minor criminal cases involving traffic tickets; there's a vast racial disparity in who gets suspended driver's licenses in states like North Carolina. Indeed, both race and poverty explain these enormously disparate patterns. At the most serious end of the criminal justice system, as I described in my book, there is an enormously racially disparate pattern in terms of who gets death penalty and who gets executed. It’s a White Lives Matter effect: cases with white victims are far, far more likely to result in a death sentence than cases with Black victims. There are dozens of individual level of studies who have confirmed this pattern. We saw similar disparities in juvenile “life without parole” sentences.

There are many inflection points in which police, prosecutors, judges, and jurors, each have an enormous amount of discretion. There is strong evidence of enormous racial disparities that result at each of those points. And the question is: how do you inform that discretion in a way that doesn't replicate biases? The national conversation that is happening now is really important because at some point, as a researcher, you can document these racial disparities. But what happens next? It takes a broader social and political movement to create action around longstanding racial disparities.

And what do you think is the role of scientists in such movement?

Brandon Garrett: Individual stories are incredibly compelling. But there's a danger in relying on anecdotes that can be misleading. We need systemic data to ensure that the stories that people find compelling are representative. It is important that researchers from different disciplines be involved to make sure that reforms aren't counterproductive. The leading example was perhaps the introduction of sentencing guidelines, which some thought would reduce race disparities in sentencing, as judges would have less discretion. Instead, we saw a remarkable surge in the severity of sentences across the country.

Could you tell us a little about the book that you're working on right now?

Brandon Garrett: In spring 2021, my new book “Autopsy of a Crime Lab: Exposing the Flaws in Forensics,” will be published by California Press. It walks through each of the different ways that forensic science can go wrong. Many of us assume that having science in court is a good thing. I have studied DNA exonerations. Had it not been for DNA technology, those innocent people might have spent the rest of their days in prison for crimes that they did not commit. However, in doing that work I quickly learned that forensic science also played a role in sending those innocent people to prison in the first place. In my new book, I carefully unpack one step at a time: How did this happen? Was it crime scene contamination? Was it the reliability of the forensic method itself? Was it the way that it was presented in court? Was it presented in a misleading way? Was it some type of pro-law-enforcement bias among the experts? Was it that the experts not proficient? Was it a failure of lab quality controls? Did the judges fail to keep out junk science?

The book unpacks step by step all the different ways, from the crime scene, through the lab, to the courtroom, that errors continue to occur.  The tragic results have included wrongful convictions, shut downs and audits of entire crime labs, and contaminated criminal investigations. This is not “a bad apple problem.” There are systemic problems with the use of forensic evidence in our criminal system.

Do you believe that scientific misconduct occurs because of bad apples or systemic issues? What are three important things in your view that help maintain a strong research integrity culture?

Brandon Garrett: I believe that in general, systemic issues explain why individual bad conduct can go on undetected, to create real harm.  Sound compliance systems, that include regular blind testing and auditing, are important in every domain, whether it is corporate compliance or a laboratory. A strong research integrity culture is important, but the shared values must be accompanied by routine empirical testing.