Bruce Davidson / Magnum Photos

What We're Reading

Greg Berman

September 27, 2023

Recent research. Quick takes.

Recent research. Quick takes.

This roundup of recent research covers a range of topics of enduring interest to us, including policing, violence and the challenges of government implementation. As is often the case, the research in question offers both encouragement and cautionary notes for those who care about advancing safety and justice. But let’s focus on the positive: If these studies are accurate, incarceration rates and racial disparities are going down, the criminal justice system is solving more murder cases than you might think, and microtargeted, hot-spot policing can reduce crime while minimizing community conflict. We hope researchers and practitioners will tease out the implications of these findings in the months ahead.

As always, if you have suggestions for what to include in “What We’re Reading,” please email us at

The Sixty-Year Trajectory of Homicide Clearance Rates: Toward a Better Understanding of the Great Decline

“Clearance rates,” once a bit of inside-baseball criminal justice vernacular describing the percentage of cases that have been successfully resolved, are having a moment in the sun. “Solve more crimes: NYPD clearance rates need to rise,” declares the New York Daily News editorial board. “Is the great decline in homicide clearance rates all a big misunderstanding?” ponders Matt Yglesias in his influential Substack newsletter, Slow Boring.

By looking at the history of American clearance rates, Vital City contributor Philip J. Cook and criminologist Ashley Mancik make an essential contribution to the public discourse about police performance in this provocative essay for The Annual Review of Criminology. Their specific target is the alarming finding that homicide clearance rates have been declining for more than a generation, with nearly half of all homicides going unsolved in 2020. “Are the historical trends in the national clearance-rate statistics an accurate reflection of reality?” ask Cook and Mancik. The short answer: “It’s complicated.”

Cook and Mancik highlight a disturbing shift that has taken place since the 1980s, with particularly stark declines in clearance rates for gun homicides involving Black victims. But Cook and Mancik also suggest that at the same time, there has been an “upward trend” in the quality of police investigations thanks to computerized databases, DNA analysis, video and digital evidence, and other advances in forensics. Moreover, they argue that the likelihood of conviction and punishment has actually increased in recent years — “the clearance rate declined markedly but the ratio of prison admissions to homicides actually increased between 1970 and the 1990s” — thanks to increased standards for making arrests.

While Cook and Mancik are not able to conclusively solve the “criminological mystery” of why clearance rates have declined in the U.S., their paper brings real nuance to the public debate and raises important questions about the usefulness of clearance rates as an indicator of police effectiveness.

Bottom Line: “The Great Decline — a nearly 30 percentage point drop in the homicide clearance rate between the 1960s and the 1990s … is most likely true. Although it is hard to believe that there was a period in many of our lifetimes when the homicide clearance rate was over 90%, that appears to have been the case … But our analysis strongly suggests that the historical pattern of the ratio of prison admissions to homicides trended upward during much of the period of the Great Decline. If true, the implication is that the decline in arrests concealed an increase in what might be considered “good” arrests — those that resulted in a successful prosecution and sanction. Taking that finding at face value obviously leads to a much more positive view of this history. With a better outcome measure, we conclude that police investigations of homicide cases were increasingly successful between 1970 and 2000 — just as we might expect given increased resources and improved technology.”

A Generational Shift: Race and the Declining Lifetime Risk of Imprisonment

In recent years, a spirit of pessimism has pervaded conversations about Black progress in America, with The New York Times reporting that “many measures of Black achievement in the U.S. have stalled or reversed.” Black incarceration rates have been a particular source of concern, with prominent activists and nonprofits suggesting that 1 in 3 Black men born in 2001 will go to prison in their lifetimes.

In Demography, James P. Robey, Michael Massoglia and Michael T. Light argue that these estimates, and the pessimistic narratives they have buttressed, are no longer accurate. Using data from the National Institutes of Health, the Bureau of Justice Statistics and the Centers for Disease Control and Prevention to estimate the risk of incarceration in state and federal prisons, the authors find that “between 1999 and 2019, the Black male incarceration rate dropped by 44% and notable declines in Black male imprisonment were evident in all 50 states …We estimate that less than 1 in 5 Black men born in 2001 will be imprisoned.” White male incarceration rates also were reduced during the study period, although the decline was less pronounced. As a result, “racial disparities in incarceration are now lower than at any point in decades, although the disparity remains quite large at 6.1:1.”

While the authors issue the appropriate cautions and caveats about their work, their findings offer a strong counter to the widespread impression that “nothing has changed.”

Bottom Line: “Assertions that 1 in 3 Black males born at the turn of the century will go to prison are no longer accurate. Rather, based on current incarceration rates, our results suggest that the correct estimate for this cohort — assuming that current rates of incarceration remain stable — is 1 in 5. As incarceration rates declined considerably year over year, so too did the lifetime risk of imprisonment. Indeed, between 1999 and 2019, the lifetime risk of incarceration declined by approximately one half for Black men and by roughly 20% for Hispanic men.”


In 2022, criminologist Scott Decker received the American Society of Criminology’s Vollmer Award, named in honor of August Vollmer, the first police chief in Berkeley, California, and one of the fathers of modern American policing. Decker’s acceptance speech was recently published in Criminology & Public Policy.

According to Decker, “In this era of ‘best practices’ and ‘evidence-based programming,’ in criminal justice, we rarely hear about the modal outcome of programs and practices: Failure.” Full credit to Decker for using his turn in the spotlight to deliver this particular message to the field, however unwelcome. (Full disclosure: A few years ago, Aubrey Fox and I wrote a book with a similar lesson called “Trial & Error in Criminal Justice Reform: Learning from Failure.”)

Decker focuses on several different kinds of failure, including the failure to innovate, the failure to evaluate and the failure to sustain successful interventions over the long haul. But his primary thrust is that “many delinquency and crime programs fail at the implementation stage.” That is, too little time and attention is devoted to ensuring that programs are well designed, well managed and faithful to the underlying model. In relaying the stories of several disappointing programs that he was directly involved in evaluating, Decker highlights a recurring source of failure: Changes in leadership, particularly within police departments, are often a moment when programs are deprioritized or jettisoned altogether.

Bottom Line: “Criminologists pay too little attention to the literature on implementation science.”

National Guide to Improving Court Appearances

Ideas42, a nonprofit design firm that specializes in using behavioral science research to address thorny social problems, turns its attention to court reform. Specifically, they look at the problem of court nonappearance. This is a decidedly unsexy problem but one with major ramifications: Ideas42 estimates that missed court dates cost state and local governments tens of millions of dollars per year — to say nothing of the cost to litigants and their families, who face fines, arrest and possible jail time.

Rather than assuming that absence from court is intentional and relying on sanctions to ensure compliance, Ideas42 argues that court systems should actively encourage appearance by lowering some of the barriers — unclear information, transportation problems, lack of childcare, fear of the system, etc. — that often prevent people from showing up.

The report outlines an array of concrete suggestions for improving court appearance rates. These range from relatively small and simple enhancements (e.g., improving the design of court forms, automatically enrolling litigants for court reminders via text, calls and emails) to more ambitious and challenging system reforms like providing transportation assistance, improving access to counsel and offering social services and case management to litigants.

Bottom Line: “Improving court appearance is vital: It moves cases forward to serve justice, saves time and money for all and prevents people from getting snarled in the system. Yet we’re increasingly learning that relying solely on the traditional approach to fixing nonappearance — punishing people who miss court — simply cannot fix the problem. Court users lead complex and demanding lives; what they need to meet their obligations are the resources and supports to match those realities … We urge courts to start with the premise that court users want to meet their legal obligation to attend court … The legal system, by nature, is rooted in precedent — yet to solve a problem like court appearance, we need to adopt new methods and imagine processes to better serve the system and court users.”

Do Cops Know Who to Stop? Assessing Optimizing Models of Police Behavior with a Natural Experiment

In this working paper for the National Bureau of Economic Research, David Abrams, Hanming Fang, and Priyanka Goonetilleke wade into heated debates over race and policing:

"Following the death of George Floyd in 2020, there were calls to “defund the police” … Defunding the police, however, does not come without costs in the form of potentially increasing both violent and non-violent crimes, due to lack of deterrence and punishment. The key to responsible police reform, therefore, is to assess the following trade-offs: how much will crime increase as police activity wanes? Surprisingly, the answer to this question is both theoretically ambiguous and empirically challenging."

In an attempt to help answer this question, the authors focus on the use of stop-and-frisk in Chicago and Philadelphia. Their dataset comprises more than 2.5 million combined stops in the two cities and takes advantage of a “natural experiment” that substantially altered policing practice: “In mid-March 2020, there was a 61% and 34% decline in pedestrian police stops in Chicago and in Philadelphia respectively. After increasing back toward normal levels, there was again a dramatic decline of 44% in Chicago and 74% in Philadelphia in stops following the killing of George Floyd.”

Comparing results before and after these changes, the authors find ambiguous results. As police stops decreased in both cities, the “hit rates” — the percentage of stops where police successfully found guns or other contraband — also increased significantly. This makes sense — as police limited their activity, they likely concentrated their enforcement efforts on those engaged in the most suspicious behaviors.

While the data suggest that the use of stop-and-frisk can be an effective tool in finding contraband, the authors also issue this cautionary note: “Although stop-and-frisk policing has long been justified based on an asserted impact on crime, we find no evidence for that in our data.”

Bottom Line: “Using extremely granular data from Chicago and Philadelphia, … we find that hit rates generally increased as police stops plunged ... At the same time, the rate of legally unfounded stops fell. [This] suggests that the impact wasn’t due to a change in police effort, the composition of individuals on the street, changes in police deployment, or crime. While both the total number of police making stops and the number of stops per officer fell, the latter seems to be more responsible for the increased hit rate … No evidence for a deterrent impact of police stops and frisks is found.”

Hot spots policing as part of a city-wide violent crime reduction strategy: Initial evidence from Dallas

Our friend David Weisburd has been a prominent advocate for hot-spot policing, suggesting that police can effectively combat criminal behavior by targeting their resources on the micro-locations (a specific corner, a poorly lit alleyway, etc.) where crime tends to cluster.

This essay in the Journal of Criminal Justice by Michael R. Smith, Rob Tillyer and Brandon Tregle examines the impact of a hot-spot policing effort launched in Dallas, Texas, in 2021.

The Dallas police department divided the city into more than 100,000 grids and focused their energy on the 50 grids with the highest counts of violent incidents. (More than 92% of the grids did not experience a single violent incident during the study period.) Interventions included dispatching a patrol car to sit in high visibility areas during peak crime hours or deploying officers to identify repeat offenders and execute warrants. These efforts were supplemented by the installation of pole cameras and patrols by officers on bikes.

The research team found that the hot-spot treatment strategy resulted in a 10.7% reduction in expected violent crime incidents. Significantly, the reductions endured even after the hot-spot strategy was discontinued: “violent crime incidents decreased by 14.2% in the treated grids during the three-month period following treatment compared to non-treated grids.”

Bottom Line: “Contemporary hot spots policing strategies are effective in reducing crime in small geographic locations. Overall, our results show that treated areas experienced significant reductions in violent crime compared to areas that did not receive treatment. Importantly, these reductions in violent crime did not displace crime to surrounding areas … Finally, reductions in violent crime were not achieved via heavy handed, over policing tactics that have been shown to inflame citizen-police relationships.”

Preventing Gun Violence in America: What Works and What Is Possible

The American Academy of Political and Social Science has devoted a special issue of The ANNALS journal to the problem of gun violence. This important collection includes essays from a broad range of disciplines and authors, including pieces from Vital City contributors Joseph Richardson Jr., Philip J. Cook, Shani Buggs and Anthony A. Braga. Taken together, the contributions offer a detailed look at American gun policy and the kind of interventions that might make sense given the unique contours of the American legal and cultural environment.

In their introductory essay, co-editors Kerri M. Raissian, Jennifer Necci Dineen, and Cassandra Crifasi explain that they seek to “elevate the conversation” about gun violence in America, which they characterize as “dominated by polarized and politicized factions.”

In pursuit of this goal, they ground the special issue in national data, making the point that “gun violence is pervasive in the U.S., with gun death and injury not just persisting as a major public health crisis in the most recent decade — but worsening.” But, crucially, they also dig deeper, explaining that there is broad variation from state to state — in some places, gun violence is rising and in some places it isn’t. Even more importantly, different kinds of places have different gun violence: “Wyoming’s high suicide rate implies that gun deaths there are concentrated among older, White men; whereas Mississippi’s markedly higher gun homicide rates implies that young, Black men are the most likely victims of gun death. As such, gun death reduction strategies that work in Mississippi may not be well suited for Wyoming. In short, America does not simply have a gun epidemic; it has multiple gun epidemics that likely require a variety of different responses.”

Bottom Line: “Although guns have always been part of American life, the level of mayhem they are currently creating is unprecedented. That said, the rates of gun injury and death rates vary significantly by state, ranging from a low of 3.4 gun deaths per 100,000 in Hawaii to a high of 28.6 per 100,000 in Mississippi. In this ANNALS volume, we set out to keep those two ideas in mind at the same time. Guns can continue to be part of the culture, and we can bring firearm injury and death down from their current, crisis levels.”