Communities across the country are debating the role mobile crisis response programs can play in their emergency response. Typically, these types of programs take calls from 911 dispatchers that require assistance with mental health, substance use, welfare checks and other instances that do not require a law enforcement response. The debate continues because there hasn’t been a formal method to measure impact of these unarmed emergency-responding teams focused on deescalation, until now. University of Oregon researchers developed a new way for communities to measure if and how much their programs divert calls from police.
In a study published Sept. 23 in Science Advances , the researchers outline a way to measure how much demand these programs divert from police. Using data from Eugene, Ore.’s, long-running CAHOOTS program, they developed a measurement standard that found the program’s impact was substantially greater than some previous estimates suggested.
Rori Rohlfs, an associate professor of data science; Claire Herbert, an associate professor of sociology; and Nathan Burton, a UO alum now pursuing a master’s degree at Johns Hopkins University, developed three standardized measures: the Substitution Diversion Rate, Prevention-adjusted Diversion Rate, and Overlapping-mandate Diversion Rate. Together, the measures give communities different ways to evaluate how mobile crisis programs affect demand for police services within their emergency response systems.
“Until now, there has been a lot of talk about programs’ ‘diversion rates’ without a clear, consistent definition,” Rohlfs said. “Our research defines what we mean by diversion and gives communities different ways to measure a program’s impact depending on the question they are trying to answer.”
Simply comparing call counts handled by police and crisis responders can be misleading because the two services don’t necessarily respond to the same kinds of calls. Programs also vary widely in their staffing, responsibilities and the types of calls they are equipped to handle.
The Substitution Diversion Rate measures mobile crisis program responses that directly replace police response; the Prevention-adjusted Diversion Rate additionally includes reduction in police responses likely due to prevention by early intervention by the mobile crisis program; and the Overlapping-mandate Diversion Rate looks specifically at calls that either police or the crisis program could handle.
Applying the new measures to CAHOOTS, the researchers estimate CAHOOTS directly diverted 18% of calls that otherwise would have gone to Eugene police. When they accounted for calls that may have been prevented from entering the emergency response system because CAHOOTS was available, the estimated impact rose to 23%. CAHOOTS also handled nearly 50% of the calls that fell within the overlapping responsibilities of police and the crisis response program.
The researchers aren’t proposing a single benchmark for what a successful mobile crisis program should look like. Instead, their framework gives communities a consistent way to measure how their own programs fit within their emergency response systems.
“Mobile crisis programs across the country operate very differently, including who responds and what kinds of calls they are equipped to handle,” Herbert said. “These measures give communities a way to understand what their particular program is doing and how it affects the larger emergency response system.”
CAHOOTS, or Crisis Assistance Helping Out on the Streets, began serving Eugene in 1989, making it the longest-running mobile crisis response program in the United States. Its two-person teams pair a crisis worker with a medic to respond to calls involving issues such as mental health crises, substance use and welfare checks, all of which are examples of 911 calls that do not require a traditional law enforcement response. The City of Eugene ceased funding for CAHOOTS in Eugene in 2025, but it does still operate in Springfield, a neighboring city.
The researchers say reliable measures can give communities better evidence for making decisions about how to allocate resources across their emergency response systems — information Herbert said would have been valuable for Eugene to have years earlier.
“As a former CAHOOTS worker who experienced the loss of the service in a very visceral way, and as a current advocate for the return of community response to the community of Eugene, I will fail to adequately express the importance of bringing receipts to this conversation,” said Laurel Lisovskis, LCSW, clinical director, crisis worker, Willamette Valley Crisis Care . “We see this academically rigorous, thoughtful work that these researchers have done to capture the weight of this loss as incredibly meaningful to our community to try and gain back a way to truly check on our neighbors and loved ones again.”
Science Advances
Quantifying CAHOOTS: Measuring a mobile crisis response program’s impact on police response through call diversion and prevention
23-Sep-2026