The recent uptick in school shootings has resulted in increased police presence in public schools. Now, school districts across America face the task of clearly defining which scenarios warrant police intervention, and which everyday challenges would be better left to teachers or other qualified administrators.

Just before Thanksgiving, the Colorado chapter of the American Civil Liberties Union settled a lawsuit with the Douglas County School District dealing with the blatant mistreatment of an 11-year-old student with autism. School resource officers arrested and detained the student for a minor instance of misconduct that was already being addressed—the upset student had poked another child with a pencil, but was calming down with the school psychologist when he was arrested. As a result of the suit, school resource officers (SROs) in the district are now required to go through more intensive training regarding interactions with students with learning or behavioral disabilities.

This unfortunate episode holds an important lesson for school districts across the country. Yes, school safety is essential, and students need to be secure from threats internal and external. This is particularly true when it comes to protecting students in real time from grave dangers like active shooter situations. But unnecessary use of law enforcement can be extremely counterproductive in other, more regular contexts.

SROs are sworn law enforcement officers who are trained and empowered to exercise authority within school settings. There were an estimated 46,000 SROs in America’s schools in 2020. Amid rising concerns over shootings in schools, some districts have also been mandating more intense searching and monitoring of students.

However, the presence of armed officers in schools has become a point of tension and debate. Critics raise concerns about the excessive use of force against students themselves, many of whom are not immediate threats to staff or other students and are simply exhibiting nonviolent delinquent behavior.

Just as policymakers and many others have been critical of some standard law enforcement practices, the community should pay equally close attention to officer practices in public schools. Interaction with school resource officers can lead to full-on criminal justice system involvement for juveniles. System involvement can have lifelong ramifications—worse for juveniles—that should be avoided if at all possible in the interest of public safety.

How then, can we ensure that what occurred in Douglas County is less likely to happen again in the future, and our nation’s schoolchildren are kept safe from all types of harm? Recommendations from advocates across the political spectrum range from full removal of law enforcement presence from schools altogether, to arming teachers with firearms. What has proven to work time and time again in standard law enforcement, however, is community policing and co-responder models.

In light of this debate, several school districts have developed positive frameworks for officer-student interactions and measured enforcement of school rules. These resemble co-responder models in standard law enforcement, emphasizing community policing, availability of clinicians or social workers who can effectively deal with mental health issues, and proactive engagement to create trust and identify points of intervention before anything dangerous occurs.

For example, the Round Rock, Texas, Independent School District has seen great success by deeply integrating its SRO program with behavioral health services. Together these provide a more holistic approach to school safety geared toward intervention before a student gets into trouble. This model helps alleviate the valid concerns of SRO critics, who point to the counterproductive nature of overcriminalizing juveniles and exposing them to a cycle that disengages them from their education to the detriment of their future success in learning and in life.

Promisingly, Philadelphia—which is markedly demographically different than Round Rock, Texas—has seen similar successes with its own shift towards “school safety officers” who focus on mentoring and supporting students, moving away from arrests and towards trauma-informed practices and restorative justice to head off potential problems in schools. One jurisdiction in Johnstown, Colorado, has gone further, formally establishing a full SRO and clinician co-responder model nearly two years ago.

Another innovative approach to SROs is the Safety Net Collaborative in Cambridge, Massachusetts, which is intended to break down silos between law enforcement, education, child welfare, and behavioral health systems. Safety Net officers are called “youth resource officers” (YROs) instead of SROs, because their goal is to build meaningful relationships with youth that extend outside of school into the wider community. Since YROs began spearheading deflection efforts in Cambridge, juvenile arrests and recidivism have declined at a statistically significant rate compared to local and national averages.

Data suggest that students can and often do build positive relationships with SROs the more they are regularly exposed to each other. This finding aligns with research seen in policing outside of schools. Co-responder models and community policing have already garnered support in the larger policing community as effective ways to handle scenarios that are not always ideally handled by “traditional” policing alone. Together, they are able to reduce the amount of misused officer time and resources, as well as reduce arrests—especially those related to mental or behavioral health.

As an Arizona state senator recently pointed out, many jurisdictions are unable to provide the resources necessary to implement a more innovative SRO program. In some cases, though, all that may be required is more positive collaboration, as in Douglas County, where the case that resulted in a lawsuit was already being effectively handled by a school psychologist.

Regardless of the current state of SROs in any jurisdiction, steps should not only be taken to ensure students remain protected from threats, but that students’ guardians aren’t actually making the situation worse. Steps towards collaboration in the style of community policing and co-responder models offer this promise.