Does It Work?

The evidence base for transit safety ambassadors is uneven. Some outcomes — lives saved through naloxone administration and CPR — are directly countable. Others — crimes deterred through visible presence, escalations prevented through de-escalation — are harder to measure. Most documented data is program-reported rather than independently evaluated. And the evidence comes almost entirely from large urban transit systems in cities with established transit infrastructure.

The life-saving outcomes are directly documented and specific. The crime reduction and ridership recovery outcomes are promising but based on limited, self-reported data from programs operating in specific contexts.

Life-Saving Outcomes

Los Angeles Metro’s official board records show that ambassadors saved 334 lives through Narcan administration and CPR as of July 2025. This figure comes from Metro’s internal tracking and has not been audited case-by-case by an independent party. [1]

A December 2025 evaluation by UCLA’s Institute of Transportation Studies reviewed the ambassador pilot, recommending the program as a potential national model. Researchers found that “safety perceptions increased over the period ambassadors were deployed” and that the program achieved “many of its initial goals.” [2]

Ambassador David Moreland, a Vietnam War veteran who served as a military medic before joining LA Metro, reported personally resuscitating five people — three through Narcan, two through CPR. [1]

Rider Perception of Safety

A Safer Cities rider survey found 63% of LA Metro riders who had seen ambassadors reported feeling safer when they see them on the system. This is a program-commissioned survey result and measures rider experience rather than underlying crime rates. [5]

The LA Metro board drew this connection explicitly, voting to make the program permanent after finding that ambassadors had “improved public safety and helped increase ridership on its transit system.” [3]

Rider perception findings should not be treated as evidence of crime reduction. The evidence on whether ambassador programs reduce crime is addressed below.

Crime Reduction: Promising but Limited Evidence

The most specific transit crime reduction data comes from Bay Area Rapid Transit’s (BART) Embarcadero and Montgomery stations in San Francisco, where a pilot running from late July through December 2025 was associated with a 53% drop in safety-related 911 calls and a 67% drop in calls tied to violent incidents including fights, assaults, and robberies. Additional data from the pilot: 82% of surveyed riders felt safer with ambassadors present, San Francisco Police Department response minutes dropped 58%, and San Francisco Fire Department response minutes dropped 23%. A critical caveat applies: the pilot period coincided with a citywide 25.8% overall crime decrease, which means the reductions at these two stations cannot be attributed exclusively to ambassador presence. These figures are program-reported from a single pilot at two stations. The attribution question is genuinely unsettled. [4]

Carnegie Mellon criminologist Dan Nagin, in “Deterrence: A Review of the Evidence by a Criminologist for Economists,” Annual Review of Economics, Vol. 5, 2013, pp. 83–105, found that visibility of authority figures “makes crime less attractive” and may be “more important than even the power of apprehension.” Nagin’s research applies to visible authority presence generally, not specifically to transit ambassadors. [7]

A randomized controlled trial of unarmed uniformed security presence in a transit system was published in PLOS ONE in 2017 by Ariel, Bland, and Sutherland (University of Cambridge and RAND Europe). Conducted in South West England’s rail stations, not in the United States: stations were randomly assigned to treatment (directed patrol by uniformed, unarmed private security agents) and control conditions over six months. The treatment produced 41% more patrol visits and 29% more patrol minutes at treated stations, associated with a 16% reduction in victim-generated crimes across station complexes and a 49% increase in police-generated detections. [Ariel, Bland & Sutherland, PLOS ONE 12(12): e0187392, 2017: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0187392]

A rigorous independent evaluation of U.S. ambassador programs and crime rates, controlling for confounding factors, has not been published for any major program. [2]

Evidence on 911 Call Reduction and Police Resource Impact

Washington D.C. Metro’s deployment of ambassadors was specifically positioned to “tamp down on crime spikes and boost security without hiring or deploying more police officers.” [WMATA Metro Ambassadors program page] LA Metro’s mission explicitly states the goal to “reserve law enforcement and armed responses to those incidents that truly warrant it.” [3]

The field does not yet have published studies measuring what percentage of transit calls are resolved by ambassadors versus requiring police escalation across multiple programs. [2]

What the UCLA Evaluation Found: Fuller Findings

StreetsBlog LA (Joe Linton) reported on UCLA’s Institute of Transportation Studies evaluation of the LA Metro ambassador pilot. Researchers found that ambassadors “advance a community safety approach towards meeting riders’ needs,” “make a positive contribution to the system,” and “support riders and operator safety and connecting vulnerable riders to resources.” [StreetsBlog LA, December 2025; UCLA ITS December 2025: https://www.its.ucla.edu/publication/la-metro-transit-ambassador-shows-promise/]

Specifically on life-saving: researchers documented that the team “assists with the first level of homelessness response, with crisis de-escalation, and by administering Narcan to prevent overdoses” and has “saved hundreds of lives on the system through Narcan use, CPR, and first aid.” [UCLA ITS, December 2025]

On visibility: researchers noted the team provides “more eyes on the system and offer a highly visible presence to riders.” [UCLA ITS, December 2025]

On program policy: the board’s decision to make the program permanent reflects “evidence that the pilot program was able to achieve many of its initial goals.” [UCLA ITS, December 2025]

Service Connection Outcomes: What Programs Document

BART’s program documents service connections to “social services and mental health nonprofits sprinkled throughout BART’s five-county service area” as a core outcome, but does not report specific service connection numbers. [9] Minneapolis’s Transit Rider Investment Program was designed in part to “connect people experiencing homelessness and addiction with social services agencies.” [Governing, Jared Brey, December 14, 2023]

The service connection evidence across transit ambassador programs consists primarily of program descriptions of activities, without documented outcome data on how many people were connected to services, how many accepted them, or what happened downstream. [2, 9]

An exception is LA Metro’s housing connection data reported by Mass Transit Magazine: the Metro team “connected 2,709 people to interim or permanent housing, exceeding the agency’s goal… by more than 150 percent” in one year. Since the program launched in 2023, the team has “helped over 645,000 people.” [Mass Transit Magazine, 2024] Mass Transit Magazine characterized the “outreach model” as “driving the results” in a “37 to 39 percent” year-to-year decrease in homelessness across the Metro system. [Mass Transit Magazine, 2024]

These figures are program-reported via Mass Transit Magazine and have not been independently verified. They represent the most specific service connection outcome data in the public record for any transit ambassador program.

What Independent Evaluations Have Found

UCLA evaluation of LA Metro (December 2025). UCLA’s Institute of Transportation Studies published an evaluation of the LA Metro ambassador pilot in December 2025. Researchers found that rider safety perceptions increased over the period ambassadors were deployed, that the program achieved many of its initial goals, and that the transition to in-house employment under the July 2025 Teamsters agreement addressed documented contractor-era problems including below-living-wage pay, high turnover, lack of break facilities, and arbitrary assignments. The evaluation also identified gaps: the need for stronger data collection systems, more comprehensive outcome tracking, and expanded bus coverage. It does not constitute a controlled study of crime reduction, and its findings are specific to LA Metro’s context. [2]

BART system-wide crime data (2025). BART reported that the full-year 2025 data shows overall crime down 41%, violent crime down 31%, property crime down 43%, robberies down 60%, auto thefts down 50%; aggravated assaults up 12%. This is a system-wide figure, not a controlled study of the ambassador program’s contribution specifically. Chief Kevin Franklin credited the combination of Crisis Intervention Specialists and other non-sworn personnel alongside sworn officers as part of the safety strategy. BART spokesperson Alicia Trost confirmed the agency has no analytical reports specifically evaluating the ambassador program — “given that it is now one of the oldest in the country.” Crime drops are attributed to multiple factors including doubled officer presence, 715 new fare gates at all 50 stations, 4,000+ surveillance cameras, and LED lighting. [8, Antioch Herald, January 2026: https://antiochherald.com/2026/01/crime-on-bart-drops-41-in-2025/; Oaklandside, February 3, 2026]

BART diversion data (Q3 2025). BART’s quarterly performance report (December 4, 2025) documents that police dispatch deployed a Crisis Intervention Specialist 206 times in the third quarter of 2025; 2,591 calls were diverted — handled by specialists instead of police. [BART Quarterly Performance Report, December 4, 2025, cited in Oaklandside, February 3, 2026]

MBTA — Massachusetts Inspector General audit (July 2023). The Massachusetts OIG found MBTA overpaid its ambassador contractor Block by Block by more than $5.3 million. The MBTA “did not set clear goals or performance metrics” to track ambassador performance and exercised insufficient contract oversight. Inspector General Jeffrey Shapiro found MBTA did not verify how ambassadors conducted elevator checks. This is the only published government audit of a U.S. transit ambassador program. [Boston Herald, July 27, 2023: https://www.bostonherald.com/2023/07/27/mbta-overpaid-for-transit-ambassadors-did-not-set-performance-metrics-report-finds/; Mass.gov: https://www.mass.gov/news/the-oigs-internal-special-audit-unit-isau-review-of-the-mbtas-in-station-customer-service-contract-with-block-by-block]

What the Evidence Does Not Show

There is no published randomized controlled trial or quasi-experimental study comparing transit crime rates before and after ambassador deployment while controlling for confounding factors. The San Francisco 911 call reduction comes from a single pilot without a control condition. The LA Metro life-saving data is program-reported without independent verification of individual cases. [2, 4]

Most programs have been operating for too short a period to generate the longitudinal data needed to assess longer-term impacts on ridership, crime, and social service utilization. The programs documenting outcomes at all — primarily LA Metro and BART — are among the largest and most resourced. Programs in smaller systems or earlier stages have little or no published outcome data. [2]

The evidence is concentrated in large urban transit systems. Los Angeles, San Francisco, and Washington D.C. are not representative of American transit systems broadly. [2]

The evidence base is sufficient to justify continued investment and to inform implementation decisions; it is not sufficient to make claims that go beyond what the data supports. [2]