After spending a year researching police practices during protests and focusing in particular on protests in New York City over the past decade, I noticed that police are confronted with two very important, but sometimes conflicting, goals when it comes to protest policing. As seen in recent protests across the country, including a series of marches in New York City to mark the anniversary of Eric Garner’s death, police are charged with preserving public safety and facilitating demonstrators’ ability to communicate their messages. Recognizing the extent to which police control the social and democratic benefits of protest activity, it is important for police departments to balance these goals.

In a recent issue of the Harvard Law and Policy Review, I suggested an approach that police departments can follow for large-scale protest events that unfold over a period of time, which allows for communication and planning between police and protesters. This approach—which is less useful when applied to flash protest events—makes it possible for law enforcement to categorize a protest, or groups within a protest, and respond appropriately during an event. Due to the variety of groups in a given march, this framework provides a way to tailor approaches to each.

In addition to achieving public safety, this approach aims to create opportunities for protesters to demonstrate solidarity, build relationships with community members, air social grievances, experiment with new ideas and cultural norms, and influence public discourse.

The model’s linchpin is police agencies identifying whether groups have explicitly disavowed violence and then tailoring their responses accordingly. Protests are often organized online in advance, giving police time to identify participating groups and review their mission statements or other organizing communications. Organizers often explicitly mention an adherence to nonviolence in their public communications.

The process is feasible because police currently monitor social media surrounding protest events and can connect groups marching under banners to statements made online or elsewhere; many people will march under signs that establish a connection to nonviolent groups. Connecting people on the ground to statements online is harder if groups march without banners, but official organizing statements—as opposed to individual posts disconnected from the central organization of the group—can usually be taken as a good indicator of why most people are there to protest.

If police see that the groups involved in a planned event reject violence, the model’s structure points to an approach of “responsive engagement.” Engagement involves communicating with protesters about ways they can address public safety concerns; responsiveness entails targeting interventions to violent individuals instead of preemptively assuming that certain groups will be violent. (See chart below.) This structure is designed to ensure that police officers treat protesters respectfully, regardless of the views they express, and that any intervention that potentially harms protesters corresponds to the genuine risk of public harm.

I think that this model creates the potential for a virtuous cycle. With more accommodating and predictable police behavior, protesters are less likely engage in violence; and with a lower likelihood of protester violence, police officers are less likely to engage in aggressive tactics. This creates an incentive for protest organizers to communicate their nonviolent intentions as a way to influence the tactics police use. A caveat, however, is that police departments must take expressed intentions of protesters at face value. For those without expressed intentions concerning the use of violence, police have reason to be on higher alert, but should continue to treat individuals as nonviolent until proven otherwise. Most protesters in this category do not intend violence and treating protesters generously demonstrates respect for the positive values protesters advance. In addition, the assumption of nonviolence avoids harmful cycles of suspicion and escalation between police and protesters.

As police departments absorb this approach, there will be new opportunities for good faith engagement between police and protesters, moving protest planning into a space of collaboration, instead of opposition, that better serves the public’s interest in a vibrant and peaceful civil society.