“Violence interrupters” are not a panacea for violent crime

As city leaders look toward alternatives to traditional policing, one approach that has gained attention is the use of “violence interrupters.” Violence interrupters are people recruited from local communities, particularly those who have a history with gangs or violence, who act as outreach workers and nontraditional caseworkers in areas where crime is occurring. These outreach workers build relationships with individuals who are at-risk for gun violence and work to support healing and address conflict through nonviolent means (e.g., conflict mediation). Some programs might help individuals find housing, education, or employment opportunities. The overarching goal is to change community norms about violence and provide pathways out of violent behavior.

The concept has gathered mainstream support among policymakers, the news media, and politicians alongside the “defund the police” movement. Even the Biden administration has described the approach as an “evidence-based” model. But is it really evidence-based?

In the short term, violence interruption programs seek to reduce shootings and murders. In the long term, they seek to reduce violent crime more broadly by shifting norms within communities. But reviews of the evidence show that the effect on shootings and murders is not consistent, and that some programs may even be associated with more violence. While some studies have found positive effects, they are few and far between and suffer from methodological flaws. The approaches might be promising in some respects, though the overall empirical evidence supporting violence interruption programs is fairly weak.

Regardless, people tout violence interruption programs as if they are a one-size-fits-all solution for violent crime. For example, a recent Time Magazine article entitled “Violent Crime in the U.S. is Surging. But We Know What to Do About It,” has been gaining widespread attention across news outlets. The piece makes the case that violence interruption programs are the key to addressing the violent crime spike, citing a few studies in support of the claim. The article emphasizes the positive findings from studies — but the same studies also often found negative or null findings. Further, the studies are not all peer-reviewed and some of the results are too preliminary to be definitive.

Violence interruption programs might work for some individuals, but the success of violence interruption initiatives strongly relies on selecting the “right” participants (which is often done on a case-by-case basis). In other words, the participants are not randomly selected for interventions, so the studies are often based on biased samples. Further, the outcomes examined are inconsistent and are not always based on crime data specifically. For example, some studies rely on surveys to evaluate peoples’ propensity toward violent behavior rather than examining crime data. Taken together, results from the studies indicate that programs might help some individuals on a case-by-case basis, but it is less clear whether these programs are successful at reducing violent crime or shootings on a broader scale.

Even among the studies showing positive impacts of violence interruption programs, researchers are unable to clearly delineate whether impacts were actually generated by interrupters themselves or whether they were associated with other factors. For example, many of the positive impacts occurred in areas where police still existed, and it is possible that some of the effects are explained via the overlap with law enforcement surveillance efforts.

Another point worth noting is that studies do not always examine the impact of programs on homicides specifically. This seems like a particularly concerning limitation, given that homicide is the most serious of violent crimes and also is a large driver of the rise in violent crime seen in 2020 and 2021. Despite the ‘promising’ results emphasized in the Time article, even cities with violence interruption programs have seen increases in homicides.

First, let’s take a look the Los Angeles Community Safety Partnership (CSP). As part of the CSP, officers work in collaboration with community stakeholders such as public health professionals, community advocates, and gang interventionists in some of the most violent neighborhoods. Officers focus connecting residents with other stakeholders and resources to help them maintain a crime-free life. In the short-term, the CSP seeks to improve public safety, and in the long-term, it seeks to generate healthier and stronger communities.

Time Magazine claims that the CSP is working to reduce crime and improve relationships with the community, based on a 2020 evaluation report. This study found that violent crime incidents decreased between 2014-2017 in target areas relative to the city as a whole, following implementation of the CSP. However, the report only shows data through 2017 and therefore does not speak to the recent violent crime increase seen in 2020 and 2021. Further, the study does not provide insights into homicide specifically, despite the fact that homicides actually increased during the time period being evaluated. Per UCR data, homicides increased from 260 to 281 between 2014 and 2017. Recent homicide counts are even higher, with 350 homicides in 2020 and approximately 399 homicides in 2021 (as of November). So, if the CSP is truly effective, why would we see a homicide increase?

Oakland, CA operates a community-oriented policing and outreach program known as Oakland Ceasefire that targets people who are highest-risk for gun violence, and focuses on connecting those people with resources they need (such as jobs, drug treatment) that can help them turn their lives around. The Time article states that the program reduced firearm homicides in the city by 31%, but this conclusion was based on a non-peer-reviewed, methodologically flawed case study.

In that study, authors showed homicide trends from 2009 to 2017 for a handful of major cities and compared them to Oakland, showing that homicides decreased during this time period relative to other major cities. However, upon closer review of the case study, it is unclear how certain cities were selected and whether the comparisons with Oakland were actually valid. Relatedly, the authors did not account for population size differences when comparing cities, which can further invalidate results. It is also important to note that the Oakland Ceasefire program emphasized elements of community-oriented policing as part of the framework, so it is quite possible that police surveillance efforts were the key mechanism in reducing crime rather than the violence interrupters. Even so, if we look toward more recent years, murders in Oakland seem to be increasing again: there were 71 murders in 2017, but this increased to 102 murders in 2020 and 123 in 2021.

Another well-known effort, Advance Peace, has been implemented in several cities throughout California, including Richmond. This program uses conflict mediation, intensive mentorship, case management, and life skills training to reach people at the highest risk for violence. In Richmond, the initiative has been regarded as successful in reducing firearm crimes by 43 percent, per the Time article. They cite a peer-reviewed study with data from 1996-2015 as authority for the assertion. This study has a decent methodology, however, results are not as clear cut as the magazine article would make you think. When taking a closer look at the results, researchers did find decreases in firearm violence, but they also found increases in non-firearm violence. Further, homicides specifically were not addressed.

In Chicago, a program known as the Rapid Employment And Development Initiative (READI) connects individuals who are at highest risk of gun violence with up to 18 months of subsidized employment and cognitive behavioral therapy, plus six months of supportive services. The Time article regards the program as effective in “substantially reducing arrests for shootings and homicides among program participants,” but they base this conclusion on very preliminary, non-peer-reviewed, non-significant results. Importantly, the experimental evaluation has not been completed and therefore final results have not been published. Further, it also remains unclear whether the program had city-wide impacts, as the program only focuses on a very specific subset of offenders (eligible people who meet certain risk assessment criteria), which could bias the sample. Another thing that doesn’t make sense is that Chicago is also seeing increases in homicides despite implementation of the READI. According to police department data, homicides increased dramatically from 2019 to 2020 (from 496 to 770), and further increased to 797 in 2021.

Cure Violence, formerly Ceasefire, is one of the most well-known violence interruption programs and it has been implemented widely. Similar to the programs already mentioned, Cure Violence uses community-based outreach workers to mediate potentially violent conflicts and connect individuals with resources. It is often touted for its positive impacts, but again, the research is incomplete and mixed. The Time article regards it as effective, based on a NYC study showing that it reduced gun injuries by 50% in one neighborhood and 37% in another neighborhood. But when you look at the actual study, reductions in gun injuries were also seen in comparison neighborhoods, meaning that reductions may not have been due exclusively to the Cure Violence program itself. Further, the study does not provide any insights into homicides.

A 2015 review examined five studies of Cure Violence programs across several American cities and found the overall quality of the evidence to be inadequate. None of the studies had consistently positive results, and even when there were positive results, they were not always statistically significant. For example, a 2009 evaluation in Chicago claimed a “41% to 73% reduction in shootings,” but this was only seen in four of the seven areas and there was no statistically significant differences between treatment and control groups. Additionally, one program in Pittsburgh was actually linked to increases in gun assaults in some neighborhoods. A 2020 review examining alternatives to police also concluded that the evidence base for violence interruption programs was “promising but mixed,” often yielding inconsistent results across sites.

Some of the inconsistent findings might be explained by the fact that violence interruption programs face inconsistent funding and support, thus, the degree to which they are successfully implemented can vary. The research also consists of largely correlational work, with no randomized controlled trials. So instead of randomly selecting neighborhoods for violent interruption programs and seeing how those places compare to similar areas, the studies typically looked at correlations between places where interrupters were believed to be active and places where interrupters were believed to be inactive. The latter is a weaker method and is more likely to find results that are related to unmeasured factors.

When considering the impact that police have on crime though, research finds strong impacts. Reviews of the research (including randomized controlled trials) have found strong evidence that specific strategies, such as hot spots policing and problem-oriented policing, are effective in reducing crime and disorder. Police strategies also appear to be better for short-term crime reduction, as officers can reduce crime in an area as soon as they are deployed there. But when it comes to violence interrupters, it is not clear if they can provide the same short-term benefits. When it comes to long-term crime prevention, there is some evidence for the effectiveness of crime prevention through environmental design (e.g. installing more streetlights, greening vacant lots) and certain types of gun laws (e.g. child access prevention laws) in reducing violent crime.

But when it comes to the effectiveness of violence interruption programs, the research as a whole suggests these programs’ results are mixed at best. Thus, the idea of completely replacing police with violence interrupters is worrisome. This is especially risky in the context of the current murder increase.