Estimated reading time: 13 minutes
Note: This is Part 1 of our two-part series on Columbus PD’s dialogue-first protest policing model. Here is Part 2 which dives into the internal challenges, political tensions, and lessons for other agencies looking to implement a similar approach.
It is late afternoon. A crowd is growing louder by the minute. Signs are raised. Someone shouts at an officer. A protest leader steps off the curb and into the street and the crowd follows.
Now it is no longer just a rally. It is an unlawful march.
The commander watches the scene unfold on camera feeds and live updates. This is the moment every incident commander dreads. The pressure is intense: protect First Amendment rights, keep traffic moving, keep officers safe, keep protesters safe, and avoid the one decision that turns a tense moment into national headlines.
In 2020, Columbus PD responded to moments like this with mass deployments, indiscriminate munitions, and mass arrests. Over one month they deployed 1,400 officers, fired 1,370 rounds of crowd-control munitions, and arrested 147 people. A federal judge later ruled their tactics unconstitutional and banned tear gas, rubber bullets, flash bangs, kettling, and more for nonviolent protesters.
Two years later, Columbus would face protests just as volatile regarding anti-war marches, police-involved shootings, extremist groups, and even attempts to disrupt the second-largest Pride parade in the Midwest.
This time?
Over 13,000 protesters.
3 arrests.
2 minor uses of force.
This is the story of how that happened.
CPD and ESIM: From Crisis to Collaboration
In the aftermath of the 2020 protests, CPD found itself operating in a fundamentally different legal environment. Long-standing “protest control” tools (tear gas, rubber bullets, flash-bangs, kettling) were no longer available, forcing the department to rethink how it managed large-scale demonstrations.
They partnered with researchers to adapt a European model known as the Elaborated Social Identity Model (ESIM), a theory of crowd psychology that rejects the idea that crowds are irrational mobs. Instead, it argues that crowds are socially structured, guided by shared identities and norms. The way police behave reshapes those identities in real time.
Heavy-handed tactics create an “us vs. them” identity.
However, when police act fairly, proportionally, and with the goal of helping rather than controlling, crowds begin to manage their own behavior.
Out of this partnership came CPD’s Public Order and Public Safety (POPS) framework.
POPS is built on three operational pillars:
- Dialogue Unit (DU): High-discretion officers whose job is communication, problem-solving, and facilitation, not enforcement.
- Bicycle Rapid Response Teams (BRRT): Low-profile crowd and traffic management.
- Arrest and Control Teams (ACT): Staged out of sight, used only for targeted interventions.
Instead of leading with force, POPS puts legitimacy and communication at the front of protest management, reserving enforcement for moments when it is truly necessary.
The Study
What makes this work stand out is not just the findings, but how they were produced. This was not a traditional evaluation conducted at arm’s length. Researchers Clifford Stott and Shahidul Hassan were embedded alongside CPD for nearly two years, working directly with commanders and Dialogue Officers across roughly 60 protest events. In turn, Columbus PD Officers Duane Mabry, Steve Dyer, and Kolin Straub also contributed to the research and to the writing of this article.
They drew on a wide range of data sources, including:
- Body-worn camera footage
- Field observations and photographs
- Focus groups with officers
- Interviews with protest leaders
- Post-event debriefs and operational reflections
Rather than treating POPS as a finished product, the research documented how the model took shape in real time across four distinct phases:
- Initial deployments
- Justice for All marches
- Anti-war and Gaza-related protests
- Columbus Pride Parade
The result is a rare inside look at how a police department and a research team co-developed a protest policing model through live operations, learning from missteps, adapting under pressure, and refining their approach event by event.
How CPD Built the Protest Policing Model
The First Test — Drag Storytime
By late 2022, Columbus PD had begun experimenting with dialogue policing, but the concept still lived on the margins. Dialogue Officers were deployed on an ad-hoc basis and few inside the department fully understood their purpose. To distinguish them from enforcement teams, they wore light-blue vests with Columbus Police Dialogue Officer stenciled on the back and positioned them as a visible alternative to riot gear.
That test arrived in December when a holiday drag storytime event was canceled after far-right groups announced they would protest. Intelligence indicated that counter-protesters were also mobilizing, making this a high-risk confrontation event even without the original programming. When the groups arrived anyway, protesters were refusing to stay on their assigned side of the roadway, prompting calls from the command post to escalate enforcement. But through conversation, the Dialogue SGT learned that the protesters’ objective was not to force entry or disrupt operations. It was to get a photo in front of the building, an image they could circulate online as proof of their presence.
Rather than authorize a show of force, he negotiated a narrow agreement: they could step forward to take the photo, but they would not enter the church and would immediately return to their designated area. They accepted and the event was a success. There were no arrests, no uses of force, and no tactical teams deployed. Eight Dialogue Officers, in light-blue vests, managed what had been projected as a mass-arrest event.
However, the backlash came the next day. A photograph circulated of the sergeant high-fiving one of the protesters, triggering accusations that CPD was siding with extremists. Within days, the department rewrote policy to prohibit gestures or behavior that could be interpreted as allegiance to any protest group.
Yet despite the controversy, CPD did not retreat. Instead, leadership expanded the Dialogue Team, using the moment to professionalize the unit and diversify its membership. The model was soon used successfully at Not Fucking Around Coalition (NFAC) demonstrations and at the neo-Nazi Blood Tribe protest targeting an LGBTQ+ event.
Drag Storytime was the program’s first real test. It proved that dialogue could prevent disorder, but also that legitimacy-based policing carries political risks that agencies must be prepared to manage.
The Ta’Kiya Young Protests — Choosing Not to Enforce
Months later, the Dialogue Policing model faced its first large-scale test. After an unarmed, pregnant Black woman, Ta’Kiya Young, was killed by police in a neighboring jurisdiction, several hundred protesters gathered in Columbus. The rally was permitted to assemble in a city park, but the permit explicitly prohibited marching in the roadway. Everyone on scene, from protest leaders to patrol officers, understood that the crowd was likely to march anyway.
The central operational question was how the department would respond when the protesters started to march. CPD made a deliberate choice: Dialogue Officers would inform protesters that entering the roadway constituted a violation, but enforcement teams would not intervene if the crowd proceeded.
In a departure from past practice, the operational commander and tactical advisors positioned themselves inside the park rather than in a distant command post. That proximity proved pivotal. It allowed the Dialogue Sergeant to communicate in real time with them as conditions evolved, eliminating the delays and misinterpretations that could drive premature escalation.
Protest organizers, however, refused to engage with Dialogue Officers and actively encouraged others to avoid contact with police. Despite that resistance, Dialogue Officers embedded themselves within the crowd as it spilled into surrounding streets.
At one point, a patrol vehicle became trapped in the congestion. Protesters banged on its windows and roof, however the officers remained calm and the crowd moved on without further incident. Dialogue Officers reported that even while walking alongside protesters, they encountered little direct hostility.
Their assignment became twofold: maintain communication and encourage the crowd to remain on one side of the roadway to preserve some traffic flow. Protesters stated explicitly that they did not want to cooperate with police, yet they stopped short of occupying the full width of the highway, even when they had the numbers to do so.
When the march first entered the roadway, Dialogue Officers temporarily assumed responsibility for traffic management in order to keep the crowd moving and prevent vehicle–pedestrian conflicts. As the event unfolded, the bike unit transitioned into that role. In after-action reviews, Dialogue Officers emphasized that this division of labor was critical: when they were tasked with traffic control, their capacity to sustain meaningful engagement deteriorated. The experience reinforced that bike teams are better suited for managing movement and roadway safety, allowing Dialogue Officers to remain focused on dialogue, trust-building, and behavioral influence within the crowd.
The event ended peacefully and there were no arrests and uses of force. Distrust of CPD remained, but something fundamental had shifted: the crowd no longer treated Dialogue Officers as adversaries. They did not embrace them, but they allowed them to operate. In the logic of dialogue policing, that tolerance was the first measurable signal of legitimacy.
Anti-War Protests — Trust in Real Time
By October 2023, Columbus was managing an entirely different scale of activity. Over the course of a several months, CPD responded to 58 protests connected to the war in Gaza, with an estimated 13,000 participants in total. Many of the events were led by student groups and newly formed coalitions unfamiliar to Dialogue Officers, and internal activist politics often worked against engagement. Organizers designated “liaisons,” but their real purpose was to block police from contacting the movement leaders.
Despite the lack of permits, CPD continued to allow marches to proceed. Instead of forcing compliance, they built informal safety bubbles around the crowds which facilitated movement, redirected traffic, and intervened only when necessary to prevent imminent harm. Notably, there was no escalation across these events. When a motorist attempted to drive through a group of protesters, protest leaders immediately stepped in, telling their members, “We do not engage,” and the crowd self-regulated without police intervention.
Trust deepened after a diaspora leader approached a Dialogue Officer following one of the marches. He offered to share advance information about planned actions. That single relationship changed the operational picture. CPD was able to deploy lighter, staff smarter, and anticipate flashpoints rather than react to them. The leader eventually became an informal liaison for multiple groups, thanking CPD after events and even assisting with future Dialogue Officer trainings to provide the protester perspective.
The most delicate moment came during a march that converged with a major college football game. Dialogue Officers tried to redirect the crowd away from the stadium district, but the group refused and pushed through one police cordon. When they reached a second cordon protecting a heavily congested area, the incident commander intervened directly, explaining the safety risks and delivering a clear ultimatum: turn west and CPD would continue facilitating the march, or keep pushing forward and lose police support. The crowd held position briefly, then turned west and the event ended peacefully.
Internally, however, this approach generated significant friction. Many officers questioned why CPD was allowing unlawful marches at all. That tension culminated in the arrest of a Black male demonstrator at a busy intersection, an enforcement action that later triggered litigation alleging wrongful arrest. The arrest disrupted the fragile trust that had been developing among protesters and reversed some of the progress toward self-regulation. Although relationships with key diaspora leaders endured, the incident reinforced a broader reality: dialogue policing remained contested within the department itself, with officers holding sharply different views on how best to maintain public order without eroding police legitimacy.
Across all 58 protests, CPD recorded only 3 arrests and 2 minor uses of force, suggesting the POPS method was associated with substantially reduced enforcement. What these protests ultimately demonstrated was not that disorder had disappeared, but that public order no longer depended on constant enforcement to exist. Through repeated, proportional, and legitimate engagement, CPD created conditions where crowds began managing their own risk, leaders emerged who could influence behavior in real time, and the department could replace massed control with targeted facilitation.
Columbus Pride — The Payoff
By June 2024, POPS had matured. Columbus Pride, drawing nearly 750,000 people, became the ultimate test. Across the country, Pride events were being shut down by protest disruptions. CPD, researchers, and organizers spent months planning, running tabletop exercises, and even facilitating conversations between parade organizers and protest groups ahead of time.
On the day of the event, a protest group eventually blocked the parade route. Dialogue Officers stepped in with an agreement that reflected two years of relationship-building: protesters could demonstrate for ten minutes, after which they would voluntarily clear the route.
They honored the deal. The parade resumed, and most attendees never realized it had been interrupted.
That is what legitimacy looks like when it is built slowly, protest by protest, mistake by mistake, and trust by trust.
Key Lessons So Far
Even at this early stage, several patterns stand out from Columbus PD’s story:
- POPS was developed in live operations, not in a classroom. CPD and embedded researchers refined the model across nearly 60 real protest events, allowing lessons to be integrated immediately rather than waiting for a final report.
- Dialogue-first deployments changed outcomes. Replacing early enforcement with facilitation consistently reduced arrests, uses of force, and crowd hostility.
- Clear role separation made the model workable. Dialogue Officers focused on relationships, bike teams managed movement and traffic, and arrest teams were staged out of sight for rare, targeted interventions.
- Trust became an operational asset. Relationships with diaspora and protest leaders produced actionable intelligence that allowed CPD to staff lighter and prevent escalation.
- The results were measurable. Across more than 13,000 participants, CPD recorded only three arrests and two minor uses of force, demonstrating that dialogue-based protest policing can deliver safer outcomes at scale.
These lessons show how a department can begin to rethink protest policing, not as a matter of controlling space, but of shaping behavior.
In Part 2 , we dive into the challenges the Dialogue Team faced, the internal and political hurdles they overcame, and what agencies can take away when implementing a dialogue-first approach.
Enjoyed this post?
Get more research translated into plain language delivered straight to your inbox twice a week.
Subscribe below and let me know which topics you want to hear more about. It only takes a minute.
If you found this helpful, share it with your team or on your socials. It helps more people make sense of research that matters.
English