With the launch of the Duke Center for Science and Justice, Duke Law School is betting that empirical, interdisciplinary research can produce evidence-based reforms in the criminal justice system.

by Jeannie Naujeck

During the years she drove on a suspended license, Andrèa “Muffin” Hudson lived in constant fear. Each time she got in her car to run errands in Durham or head to work, she would drive well below the speed limit and stop at yellow lights to avoid getting pulled over.

But it didn’t always work. “When you don’t have a driver’s license you ‘drive nervous,’” Hudson says. “Police pick up on that and it gives them a reason to run your tag.” She racked up dozens of tickets over more than a decade, including some 60 in Durham County, and spent 10 days in jail for repeatedly driving on a suspended license, all because she couldn’t pay her fines.

A civil rights activist, Hudson shared her story at Duke Law on March 25 during a lunchtime event marking the release of a report examining drivers’ license suspensions in North Carolina co-authored by L. Neil Williams, Jr. Professor of Law Brandon Garrett and William Crozier, a postdoctoral fellow in empirical legal research. The report places Hudson, now director of the North Carolina Community Bail Fund of Durham, among more than 1.2 million residents who lost driving privileges under a state law that triggers automatic, indefinite license suspension for reasons unrelated to driving. Those reasons include failing to pay traffic fines and court costs and failing to appear in court for traffic offenses.

Noting a U.S. Supreme Court finding that driver’s license suspensions cause “inconvenience and economic hardship,” Garrett and Crozier analyzed the demographic and geographic distribution of suspensions between 2010 and 2017 using data obtained from the North Carolina Administrative Office of the Courts by the N.C. Justice Center, a research partner. A key finding: Poor and minority drivers suffer the vast majority of suspensions, receiving disproportionately more than poor white drivers. 

“The harm from license suspensions is deeply felt,” Garrett says. “In most communities, particularly outside urban areas, people cannot easily fulfill the myriad obligations of everyday life — such as going to work, school, medical appointments, day care, and even the grocery store — without driving. Employers may require employees to have valid driver’s licenses. The effects go beyond the personal and are also felt in the broader economy.”

The report, “Driven to Failure: Analysis of Failure to Appear and Pay Driver’s License Suspension Policy in North Carolina,” concludes by setting out questions for future research and describing both law and policy responses to driver’s license suspensions in other jurisdictions, including constitutional challenges, restoration efforts, dismissals of charges, and legislative attempts to restore licenses and end suspensions for non-driving related traffic offenses. 


“The harm from license suspensions is deeply felt.”

— Professor Brandon Garrett 

Photo: Professor Brandon Garrett, right, meets with research associates in the Duke Center for Science and Justice. From left: postdoctoral research fellows William Crozier and Karima Modjadidi, and Arvind Krishnamurthy, a political science PhD candidate at Duke.

Building on interdisciplinary strength

Garrett’s study of driver’s license suspensions is just one of several empirical research projects he has initiated since joining the Duke Law faculty in 2018, all aimed at informing criminal justice reform. And with the launch of the Duke Center for Science and Justice, which Garrett directs, Duke Law is poised to expand its role in the effort to craft policy solutions based on empirical evidence for other problems and inequities throughout the criminal justice system in North Carolina and beyond. The interdisciplinary center focuses on three signature areas: accuracy of evidence in criminal cases, the role of risk in criminal outcomes, and addressing a person’s treatment needs.

The center builds on the substantial body of criminal justice work already being done at the Law School, including that of the Center for Criminal Justice and Professional Responsibility, which engages in training students, lawyers, prosecutors, judges, and the general public to identify, remedy, and prevent the wrongful convictions, and the Wrongful Convictions Clinic, which has won release for eight clients, most recently Charles Ray Finch in May, and Dontae Sharpe in August, who respectively spent 43 years and 25 years incarcerated for murders they did not commit.

Kerry Abrams, the James B. Duke and Benjamin N. Duke Dean of the School of Law and professor of law, calls Duke Law “the leader in criminal justice research nationwide,” citing scholarship by Charles L. B. Lowndes Professor of Law Sara Sun Beale on prosecutorial discretion; Candace M. Carroll and Leonard B. Simon Professor of Law Lisa Kern Griffin on the relationship between narrative and factual accuracy in the courtroom; Bernard M. Fishman Professor of Law Samuel W. Buell on the conceptual structure of white collar offenses; Professor of Law and Professor of Philosophy Nita Farahany JD/MA ’04, PhD ’06 on the ethical dimensions of emerging technologies in criminal law and the use of neuroscience in criminal trials; and Associate Professor of Law Ben Grunwald on the movement of convicted officers within the law enforcement labor market.

The work of a number of Duke faculty in other disciplines intersects with criminal justice, such as Professor of Economics Patrick Bayer, who has studied the effects of racial inequality on the criminal justice system, and ITT/Terry Sanford Professor Emeritus of Public Policy Studies Philip J. Cook, a renowned scholar of crime, crime prevention, violence, firearms and crime, and the economics of crime. Garrett’s research team includes postdoctoral research fellows Crozier and Karima Modjadidi as well as faculty in medicine, undergraduate and graduate research assistants, and law students enrolled in his Criminal Justice Policy Lab.

Leading researchers from Duke’s renowned School of Medicine are closely involved in the center’s work, providing platforms for criminal justice research informed by a public health perspective. Over the past year, Drs. Marvin Swartz, Jeffrey Swanson, Michele Easter, and Allison Robertson from the Services Effectiveness Research Program in the Department of Psychiatry and Behavioral Sciences have been working with research teams at Duke Law on a series of projects at the intersection of law and medicine, behavioral health, and criminal justice. The psychiatry group has longstanding interest in the effects of legal policies and intervention on behavioral health services that divert persons with mental illness or substance use conditions out of incarceration and into community-based treatment programs, and the use of alternative, or recovery, courts such as drug courts, mental health courts, and veterans courts.

“For whom and under what conditions are these programs effective in reducing criminal justice recidivism?” asks Swartz. “What are the critical ingredients in these programs? Do and, if so, how do treatment programs need to be modified to address these populations? We are excited to undertake this work with colleagues across the university.”

The answers to researchers’ questions will enhance accountability for community-based programs that seek to create better outcomes for offenders, break the cycle of recidivism, and alleviate jail overcrowding and the social problems that accompany mass incarceration, including loss of jobs and housing and the breaking up of families, while posing minimal risk to public safety. Indeed, efficacy data is increasingly becoming a requirement for the implementation of new reforms: For example, in Mecklenburg County’s pretrial risk assessment pilot program, funded by the Laura and John Arnold Foundation, eligibility restrictions were required to be supported by research. That program, started in 2014, reduced the jail population with no significant accompanying impact on public safety or increase in failures to appear in court, according to county reports.

“Mass incarceration is not only a justice issue, but a costly public health problem. It can have long-term ill effects on the physical and mental well-being of some of our most vulnerable and marginalized populations,” Swanson says. “Building evidence for interventions, public policies, and legal reforms to mitigate these consequences — or ideally to prevent criminal justice involvement in the first place — is an interdisciplinary challenge, and an exciting opportunity for health services researchers and legal scholars working together in this center.”


Building toward statewide solutions

The March release of the report on driver’s license suspensions helped build momentum for bills introduced in the North Carolina General Assembly last term, including a “Second Chance” law that would end automatic suspensions for failure to pay traffic fines. Of the Senate bill’s 14 sponsors, five were Republicans, including the Senate majority whip and the co-chair of the Senate Judiciary Committee.

“We are trying to make this a bipartisan or non-partisan effort and movement, and so having Brandon come in with his credibility and platform was a huge benefit,” says Daniel Bowes T’07, senior attorney at the N.C. Justice Center, which has long advocated for legislators to remedy the problem of rampant suspensions that can stem from such small infractions as cracked tail lights. Bowes says analysis of the data shows that using suspension as a coercive tactic doesn’t work.

N.C. Rep. Marcia Morey, a Democrat, filed bills in the state House of Representatives during the past session that would ease expunction proceedings and cap the length of suspensions. “This movement is getting the attention of all sides of the aisle and I hope we can get some positive legislation passed through,” she says.

Follow-up projects at Duke Law include surveying individuals who had their licenses suspended on the collateral consequences it had on their lives, including effects on mobility, employment, family relationships, health, and housing. Garrett’s research team hopes in the future to conduct detailed interviews with respondents, partnering with Professor Sara Sternberg Greene, who specializes in large-scale qualitative research on poverty, to better understand the impact of fines and fees on individuals and communities. The project may be expanded to other states to help build on momentum to end automatic license suspension policies; similar license suspension practices have ended recently in Virginia and Texas.

On July 1, Virginia Gov. Ralph Northam restored driving privileges for at least a year for residents who had their licenses suspended for unpaid court fees and fines; the Legal Aid Justice Center estimates more than 940,000 people had been affected. And on Sept. 1, more than 600,000 Texas drivers became eligible to have their licenses reinstated after Gov. Greg Abbott signed into law a bill ending the state’s Driver Responsibility Program, which added surcharges that could run into the thousands of dollars to traffic ticket fees, resulting in license suspensions for those who did not pay in a timely manner.

Locally, Garrett has forged a strong relationship with the Durham County District Attorney’s Office, led by Satana Deberry ’94, who took office in January 2019 with an agenda for progressive criminal justice reform, including driver’s license restoration. He and his team regularly meet with members of her office to discuss policies and research on such topics as police practices, jury selection and plea bargaining, the expansion of diversion opportunities, and pretrial decision-making. They continue to explore methods and outside resources that could make such new initiatives possible.

“Brandon is very smart about data and the power of data,” says Alyson Grine, assistant district attorney and Deberry’s team lead for policy and training. “At the outset of this administration, we sought to put in place a process for collecting data so that in a year or two we can measure the new initiatives we are putting into place and how it’s impacting the criminal justice system and the community. He’s got a skill set that is really important for us in these efforts and he’s been very generous with his time.”

As part of her effort to transform the office, in May Deberry announced new pretrial policies that include ending cash bail for most misdemeanors and low-level felonies except those involving domestic violence or physical harm. In fact, Garrett’s team had begun collecting data on pretrial outcomes over a year ago, before those policies took effect. The Duke group is collecting and analyzing data from the Durham County jail to check for disparities in how bail is ordered by different judges and identify factors that might lead to higher or lower bail, using a “webscraping” program set up by digital resource librarian Sean Chen that automatically scans the jail site for new entries and collects and records the data. Garrett, Crozier, and Modjadidi, with Duke political science graduate student Arvind Krishnamurthy, have been analyzing these data and will soon present their findings to local stakeholders. The research team also plans to analyze data on plea-bargaining outcomes.

Garrett has also served as a connector in the Durham criminal justice community, Grine says. “He has a very collaborative style that is really helpful.”


Sentencing initiatives

Garrett’s work on sentencing reform in North Carolina has been focused on the most serious sentence short of capital punishment: life without parole (LWOP). His research in this area is novel, as there have been no prior known studies of case-level data on LWOP sentencing. Partnering with Ben Finholt, staff attorney with North Carolina Prisoner Legal Services, Garrett’s team — Modjadidi and Duke political science PhD candidate Kristen Renberg, who is now also pursuing a JD — examined 94 cases statewide in which the sentence was applied to juvenile offenders. They found that county-level preferences, like those of prosecutors, drive these most-serious juvenile sentences: counties that had imposed life without parole sentences on juveniles were highly likely to continue doing so, leading to a concentration of LWOP sentences in just a handful of counties. They also found that it is now rare for such sentences to be imposed, and most of the older juvenile LWOP sentences reviewed on appeal have been reversed in recent years.

The investigators discussed their report, titled “Juvenile Life Without Parole in North Carolina,” at a Duke Law event last February that was attended by legislators and garnered extensive coverage that has already made an impact; Finholt notes that since its release, at least one district attorney has consented to a lesser sentence when he had the option to ask for life without parole.

“Since Brandon got involved, we have seen movement in the juvenile life without parole space that I don’t think we would have gotten otherwise,” Finholt says. “Every other time they had always asked for life without parole and this was the first time they didn’t. His involvement has taken the message that we were trying to promote and gotten attention that we as litigators just didn’t have either the platform or the time to do.”

Finholt credits the publicity generated by the report and event, combined with Garrett’s credibility and backing from a diverse group of funders, with bringing awareness to an issue that he’s long pressed legislators to consider. In April, Democratic Rep. Morey filed a bill in the N.C. House of Representatives that would eliminate life without parole for juveniles and replace it with parole eligibility. The bill’s three co-sponsors are Republicans.

Kristin Parks, a longtime attorney for North Carolina death row inmates and one of the bill’s authors, says Garrett contributed valuable assistance not only with its language but also with its fiscal note, an important consideration for a legislature that adjourned Oct. 31 without approving a state budget for the fiscal year that began July 1. “It is invaluable to have someone who can crunch numbers and come up with estimates of cost savings on a bill that we think is good policy,” Parks says. “The issues that we’ve been talking about for years went nowhere, and now all of a sudden people are willing to listen.

“I appreciate all he’s done because it is really important in moving our state forward and educating people who are new to these issues in different and important ways.”

Garrett’s team is currently extending the work more broadly to some 1,600 inmates serving sentences of life without parole, comparing corrections data with data concerning death-eligible cases in North Carolina. Such a detailed quantitative analysis of LWOP sentencing has not previously been conducted. They have found that LWOP sentences were strongly associated with prior practices at the county level, suggesting that local prosecutors’ preferences matter a great deal. They have also found that LWOP sentencing is affected by the race of homicide victims in a county, and by the race of defendants in cases in which the prosecutors sought the death penalty. The researchers presented a draft article with the results in September at a new Triangle-area criminal law works-in-progress series that is hosted by the Duke Center for Science and Justice. The article is forthcoming in the Journal of Criminal Law and Criminology.


A bipartisan push for criminal justice reform

The new center launches as mass incarceration, policing, and laws that disproportionately impact the poor are in the national spotlight. With 2.2 million incarcerated adults in the United States, and more than $80 billion spent annually on jails, prisons, probation, and parole, a diverse group of stakeholders is coalescing around criminal justice reform. In December 2018 President Donald Trump signed into law the First Step Act, which increases judges’ discretion to impose shorter prison sentences and allows federal prisoners, who currently number about 180,000, to earn credits toward early release based on rehabilitative programs and their risk of reoffending. Organizations as disparate as the American Civil Liberties Union and Right on Crime had joined to lobby Congress for its passage.

Writing in Slate after the First Step Act was signed, Garrett called for sound scientific and public oversight of tools used to assess offender risk. “The statute states that an algorithm will be used to score every prisoner as minimum, low, medium, or high risk,” he wrote. “But the legislation does not say how this algorithm will be designed. … The right tools need to be used.” In September, Garrett and George Mason University law professor Megan Stevenson wrote a public comment on the risk assessment system proposed to the Department of Justice. In it they raised concerns regarding the types of risk being measured, the risk cut-offs, and the adequacy of resources for rehabilitative programming.

Garrett’s cautions are based on his previous research on the implementation of risk assessment at sentencing. In a paper co-authored with University of Virginia professor of law and psychology John Monahan, he described wide variation in its application by courts and judges in Virginia and a correlation between the availability of treatment resources in a community and judicial use of risk assessment to sentence drug and property offenders to non-jail alternatives.

Reforming long-held policies — and getting those reforms right — involves analyzing the vast troves of data collected by numerous actors in the criminal justice system, including thousands of institutions, agencies, and jurisdictions. Toward that end, grant-making organizations have ramped up funding of empirical research at academic institutions. Some of the largest funders of criminal justice research include Arnold Ventures, which has provided support for Garrett’s research on eyewitness evidence, and the Charles Koch Foundation, which supports research and educational programs in areas such as criminal justice and policing reform, free expression, foreign policy, economic opportunity, and innovation, and is providing a $4.7 million grant for the Center for Science and Justice. Garrett is also a principal investigator and on the leadership team for the Center for Statistics and Applications in Forensic Evidence (CSAFE), a collaboration funded by the National Institute for Standards and Technology. That federally-supported effort includes work to improve how forensic evidence is used in court and to introduce statistical methods to forensic labs.

“There has definitely been more funding in this area over the last couple of years,” says Grunwald, whose recent quantitative research projects include empirical studies of plea discounts and trial penalties and an analysis of the reasoning behind more than 2 million investigative stops by the New York City Police Department.

“For the past few years there has been bipartisan support for a certain kind of criminal justice reform, and it’s not all that surprising. There are lots of different reasons to think the criminal justice system needs to be fixed, and people with different ideological backgrounds don’t have to hook into all of them to recognize that something’s got to change.”


Duke increases empirical research support

Historically, criminal justice data — from crime statistics to bail and sentencing information — has been difficult to access and analyze because it is collected by numerous agencies in multiple jurisdictions, often in differing formats. 

“The criminal justice system is divided into a lot of big institutions and sometimes you need data from multiple ones to really study the question you want rigorously,” says Grunwald, who estimates he spends 85% of his time on projects processing data. 

“It feels very hard to get data. And even if you get data from one state there are still 49 other states and their data is all totally different. So it’s quite a challenge. Even when the data exists, it could take months for a researcher to access it.”

To enhance empirical scholarship, Duke Law recently hired Alex Jakubow as associate director of empirical research and data support services, a newly created position in the J. Michael Goodson Law Library. In his previous post as empirical research librarian at the University of Virginia Law School, Jakubow helped create its Legal Data Lab that houses databases including a corporate prosecution registry maintained in collaboration with Duke Law. At UVA, Jakubow also collaborated with Garrett on a paper on the decline of the death penalty in the U.S. and the paper involving the use of risk assessment tools in Virginia. At Duke, Jakubow is already working with the new center on a project tracking outcomes in district attorney elections.

Jakubow says his team will help faculty and students define their research questions in empirical terms, identify data sources, extract data and work to get it into a usable format, and assist with analyzing it. 

“There’s so much data out there but a lot of it is embedded in formats that aren’t necessarily accessible or don’t lend themselves to analysis,” he says. “The role of the library is to support empirical research during the entire life cycle of a project. Fundamentally, it’s about empowering our faculty to do what they do best, in any way that we can help with that quantitative research part.”

For the driver’s license project, Crozier spent two months cleansing and putting data from the state Administrative Office of the Courts in a format that could be analyzed. That analysis revealed correlations between race, poverty, and suspension rates that illustrate why problems like North Carolina’s high rate of license suspension require solid empirical research and a thoughtful approach to finding effective solutions. 

“If you look at the numbers, you can clearly see the racial disparity: white people are not being suspended at the same rate they should be if it’s just randomly based on the population,” Crozier says. “I think you get a really good feel for how broad this problem is and how much work needs to go into finding solutions. It shows, I think, that a solution might not be as easy as finding the poorest people in the county and giving them money to pay off their suspensions.”


Science and justice in the classroom

With rising interest in criminal justice reform, the Center for Science and Justice will further engage students at Duke Law through courses and in research outside of class in efforts to improve the system. 

This fall Garrett is collaborating with Farahany, who directs the Duke Science and Society Initiative, to teach an Amicus Lab course in which Duke Law students draft amicus briefs to educate courts on law, science, and technology issues in appellate cases in several states. 

As of mid November, the class had filed three briefs: one presenting current research on blood pattern analysis that refutes testimony used to convict a Texas man of murder in 1986 and 1989; one presenting scientific studies on genetic predisposition toward antisocial and aggressive behaviors that challenges the exclusion of such evidence in the trial of a New Mexico man convicted of second-degree murder; and one presenting research on eyewitness memory that challenges the in-court identification of a Colorado man petitioning the U.S. Supreme Court for review of his conviction.

Law students also recently participated with practicing lawyers in a three-day forensics litigation course, supported by CSAFE, in which they prepared for a mock trial and had the opportunity to cross-examine a fingerprint analyst. In the spring semester, students in Garrett’s Criminal Justice Policy Lab will hear from North Carolina judges, lawyers, and lawmakers as they craft policy proposals to improve the system.

Through additional philanthropic support, Duke hopes to expand the educational component of the Center for Science and Justice in a second phase, supporting students who are entering criminal justice careers through scholarship aid, internship funding, a criminal-justice focused curriculum, and opportunities for interdisciplinary engagement with graduate and undergraduate students. The Law School also hopes to launch a criminal justice clinic to provide training in how to litigate a criminal case at the pre-trial and trial stage. 

Abrams calls the launch of the center “extraordinarily exciting and important for Duke Law and the study and reform of criminal justice around the world.”

“There couldn’t be a better place to host a major center for the study of the role of science in criminal justice reform,” she says. “We very much hope that additional foundations and individuals will follow in contributing resources to support its mission.”

From the Fall 2019 issue of Duke Law Magazine.