Photo: silhouette of hand holding handcuffs in a jail

Photo credit/Getty Images

George Floyd’s death at the hands of Minneapolis police officers in May 2020—and the demonstrations it sparked around the world— has lent renewed urgency to calls for an overhaul of a U.S. criminal justice system that a growing majority of Americans perceive as rife with racial bias.

Seventy percent of Americans, and two-thirds of white Americans, believe that last summer’s protests are due to long-standing tensions between Black citizens and the police, according to a Pew Research Center survey. The same study found that only 9 percent of white Americans say they have been unfairly stopped by police, compared to 45 percent of Black Americans who report this. 

Other analyses have pointed to systemic racism more broadly—in pre-trial detention, jury selection, and sentencing, among other areas.

But within the growing body of research on these racial disparities, little scholarship so far has focused specifically on how race figures into how the criminal justice system handles the lowest-level crimes.

“To date there has been no research on bias in the prosecution of misdemeanors,” says Anna Harvey, a professor in the Department of Politics and director of NYU’s Public Safety Lab. “While there has been work that has found unexplained racial disparities in charging decisions made by federal prosecutors as well as unexplained racial disparities in bail decisions and in sentencing decisions, studies on which misdemeanors are prosecuted have yet to be conducted.” 

The potential for racial bias in low-level crimes is even more significant in light of a new National Bureau of Economic Research working paper by Harvey and her colleagues that captures the effects of misdemeanor prosecution on future arrest and prosecution rates. The research team—which included Texas A & M’s Jennifer Doleac and Rutgers University’s Amanda Agan—examined nonviolent misdemeanor cases in the office of the Suffolk County (Mass.) District Attorney's Office. It found that defendants prosecuted for nonviolent misdemeanors had substantially higher risks of future arrest and prosecution—including for felonies and violent offenses—than those not prosecuted. 

“In other words, prosecuting these nonviolent misdemeanor cases actually increased the likelihood of later arrests, including arrests for felony and violent offenses,” Harvey explains. 

Her next step is to determine if there is evidence of racial bias in the prosecution of misdemeanors—which she notes make up over 80 percent of the cases processed by the U.S. criminal justice system.

“In Suffolk County, 46 percent of nonviolent misdemeanor defendants are Black, but only 24 percent of the county’s population is Black,” Harvey observes. “Prosecuting nonviolent misdemeanor offenses in this county is disproportionately causing Black residents to have continued criminal justice involvement.”

With support from the William T. Grant Foundation, Harvey, along with Doleac and Agan, will study the role of prosecutors' decisions in the perpetuation of racial, ethnic, and socioeconomic disparities in the criminal justice system, with a particular focus on misdemeanor defendants between the ages of 18 and 25. The work is being conducted in partnership with district attorney offices in New York, Massachusetts, and Missouri. 

“Evidence from the project about how prosecutorial decisions shape inequalities may inform the development of alternative prosecutorial strategies and lead to policy reforms to reduce these inequalities,” observes Harvey. 

Harvey spoke with NYU News about the potential implications of this timely work—and about the role higher education can play in addressing long-standing inequities.

Your research suggests that aggressively prosecuting low-level crimes could actually lead to more crime—a finding that some may find counterintuitive. How do you explain this effect?

A growing body of evidence is revealing that prosecuting arrests may actually increase the likelihood that defendants are later rearrested for subsequent offenses. Part of this effect may work through pretrial detention: defendants who are detained pretrial have both worse subsequent labor market outcomes, and increased likelihoods of subsequent re-arrest. These negative effects of pretrial detention may derive from the consequences of defendants being separated from their jobs and families while they are detained. Another part of this effect may work through the consequences of having a criminal record: defendants with records of criminal convictions again have worse labor market outcomes and increased likelihoods of subsequent re-arrest, relative to defendants without these records. Criminal records of arrest, even when defendants were not convicted, may similarly damage individuals’ labor market outcomes. Finally, part of the effect may work through the negative effects of requiring individuals to remain engaged in lengthy prosecutions, even if an individual is not detained pretrial, and even if the case does not result in a conviction. 

Why was the Suffolk County District Attorney’s Office interested in cooperating with you on this research? 

As a candidate, Rachael Rollins, now the district attorney of Suffolk County, spoke to many criminal legal stakeholders, including defense attorneys, returning citizens, those still incarcerated, and residents of the most disproportionately prosecuted neighborhoods in the county. They all conveyed to her that prosecution and incarceration for minor crimes and misdemeanors were hindering rather than improving public safety. Upon being elected in 2018, District Attorney Rollins issued a decline-to-prosecute policy for 15 categories of minor crimes—absent compelling evidence—based on these qualitative findings. When she learned that our research team was seeking to analyze quantitative data on this question, she was eager to share her office’s data and contribute to the research in this field.

How do you envision that your work, once completed, could transform law enforcement more broadly?

There are lots of things that we don't know about what works and what doesn't in law enforcement. For a long time, jurisdictions concerned about public safety have simply pursued the most punitive policies possible, in the belief that more punitive policies—for example, more stops, more arrests, more prosecutions, longer sentences—produce more public safety. But we're starting to learn that more punitive policies don't always produce more public safety. In fact, they may make us less safe. We have found that, by reducing the frequency with which nonviolent misdemeanor arrests are prosecuted, district attorneys can reduce the frequency of subsequent arrests. Perhaps not only prosecutors’ offices, but also police departments, will dial back the frequency of nonviolent misdemeanor arrests as a consequence of our work. Likewise, if we find through our work that reducing disparities in prosecutorial decisions can help to reduce disparities in outcomes, then perhaps prosecutors’ offices will be able to move forward with policies that achieve that goal.

The killing of George Floyd has renewed our focus on the historic unfairness of the criminal justice system, and, specifically, the treatment of African Americans by police. How can the academy contribute to reform?

That’s a great question. I do believe there is an important role for the academy in bringing evidence to bear upon pressing social questions. We have so much human capital in universities in the form of our faculty and our students, and many of us want to be able to use our training to help to identify those policies that may move us towards a more just and productive society. At the same time, it’s important that we also understand our limitations. It is members of impacted communities, who bear the daily weight of living in an unjust society, who should lead efforts to reform our policies and practices. Our contribution should be simply to support their work as best we can.