New York City’s new commissioner of social services has built a reputation for using data, partnerships and evaluation to tackle the hardest social problems cities face.
Local agency appointments are often dismissed as insider news — meaningful to City Hall watchers, distant to the public. Mayor Zohran Mamdani’s appointment of Erin Dalton as commissioner of New York City’s Department of Social Services deserves a different reaction. Though the choice of Dalton was met by some with an eyeroll, as though someone from out of town, in this case from Pittsburgh, had no business taking such a big job in the nation’s largest city, it is in fact a signal that competence and evidence matter in governing a city — and that management choices in human services can translate into real, measurable changes in people’s lives.
That may sound obvious. But in an era of deep cynicism about government, it is worth saying aloud. It would have been easier for Mamdani to promote someone from within city government or reward a political ally. He looked outward across the country and recruited someone widely respected for doing something that often feels rare in public policy today: building systems that learn and use data, partnership and evaluation to improve results over time.
Dalton’s tenure as the Director of Allegheny County’s Department of Human Services (DHS) has been closely associated with a culture of experimentation, collaboration and rigorous evaluation. Again and again, when new approaches to complex social problems are discussed, Allegheny DHS shows up as a partner or model.
Developing integrated data systems that link information across housing, health, schools and human services? Allegheny DHS built one of the strongest local government models in the country — and policy decisions flow from what those data reveal. It is not surprising that Dalton led the Office of Analytics, Planning, and Information Technology at Allegheny DHS before she was promoted to oversee all of DHS. Among many awards and accolades over the years, in 2024, Allegheny County was featured by the U.S. National Institute of Justice in the opening plenary of its 2024 National Research Conference, titled “Advancing Justice Through Science: Tracing Allegheny County’s Journey.”
Pioneering innovations in housing and supportive services for frequent utilizers of public services — people repeatedly cycling through emergency rooms, homeless shelters and jails who typically have long histories of mental health issues and substance use? Allegheny DHS used its integrated data system to identify key intervention points, then created systems and strategies that promote long-term wellness, reduce repeated high-intensity service use and in turn lower costs across emergency, housing, justice and health systems. For instance, Allegheny recently became the first jurisdiction in Pennsylvania to offer assisted outpatient services for individuals with mental health and substance use disorders, adding an important service option designed to stabilize people in the community and reduce unnecessary court involvement.
Rolling out alternative crisis response? Not surprisingly, Allegheny County’s focus on reducing unnecessary public costs and system involvement led them to build and measure a system in which clinicians — not just police — respond to eligible mental health emergency calls, with police support when safety requires it. Although the program, which sends an “A-Team” to respond to nonviolent, non-medical 911 calls at the request of the participating police departments, is too new to have rigorous outcome evaluation results at this date, Allegheny DHS built a public dashboard for the system to provide the public with up-to-date information that displays call volume over time, response types, call types and response outcomes. The dashboard also supports data and statistical analyses and interactive visualizations. A quick look at the dashboard’s “interventions and outcomes” page shows that between October 15, 2024 and November 10, 2025, the A-Team was able to assess/interact with clients in 95% of responses.
Bolstering prisoner reentry support and jail-cycling reform? Allegheny DHS coordinated supervision and services beginning inside the jail and continuing into the community. An Urban Institute evaluation found a 24-percentage-point reduction in rearrest risk, which translates to roughly two dozen fewer rearrests per 100 people served.
Understanding the harms and costs of violent criminal victimization in order to help drive prevention? When our research team at Temple University and NORC at the University of Chicago launched the NIJ-funded HAVEN project, Allegheny was a natural partner. Dalton and her team shared their integrated data system to help quantify the often-hidden health and economic consequences that follow violent victimization.
Creating a community violence intervention “ecosystem” where anti-violence programs are carefully layered together to fill service gaps for both the people and places most touched by violence? Allegheny County DHS was featured as one of three “jurisdiction immersion” sites at the 2025 Cities United convening, where other jurisdictions come to learn about successes in creating and sustaining a comprehensive anti-violence strategy that includes prevention and intervention. In a country with thousands of urban jurisdictions, that kind of recognition is rare.
Getting swifter and smarter child welfare decisions after allegations of maltreatment to help predict foster care placement? Allegheny DHS experimented publicly with a machine-learning screening tool (the Allegheny Family Screening Tool or “AFST”) that pre-selects investigation as the starting choice for the highest-risk referrals. An evaluation commissioned by Allegheny County found it reduced Black-white disparities in investigation, case opening and removal rates. More specifically, the evaluation found a 20% reduction in first entries into foster care from 2016-20, the years when AFST was implemented. Additionally, a later study shows the potential power of AFST for child protection: The model’s risk scores strongly predicted serious child harm measured through hospital injury encounters. This means the algorithm is effective at identifying children most likely to experience serious physical harm, allowing agencies to prioritize resources and interventions for families facing the greatest risk.
Dalton has had a significant hand in shaping all these innovations.
I first met Erin about 25 years ago when she was at the National Institute of Justice as a project manager for the Strategic Approaches to Community Safety Initiative. What stood out even then was how consistently she emphasized two things she viewed as nonnegotiable: partnerships and good data.
Those are also the two hardest things to achieve in complex social policy. Partnerships require agencies with different missions and incentives to share responsibility and information. Fealty to data requires careful and ethical governance — and the discipline to measure outcomes honestly, even when the evidence forces you to rethink what you are doing. And not surprisingly, Allegheny DHS carefully surveys clients, community members and providers — asking them how to best meet community needs and where they could do better. They use the feedback to add and change programs, shift funding and improve program monitoring.
Allegheny County DHS repeatedly invites evaluation. It partners with researchers, publishes dashboards and allows (usually by directly asking) outside researchers to test whether programs actually worked. In many jurisdictions across the country, evaluation is treated as a threat. In Allegheny, it became part of governance.
Dalton did not build these systems alone. But she clearly created a culture where experimentation was encouraged, collaboration expected and data used to guide decisions.
That matters in New York.
New York’s Department of Social Services, with a budget of $18.5 billion and roughly 12,500 employees, sits at the center of housing stability, poverty reduction and crisis response in the largest city in the United States. It oversees the city’s core public benefits operation and its homelessness response system — administering major assistance and rental subsidy programs and managing the shelter and outreach network delivered through a vast nonprofit provider infrastructure. Yet many of the systems most closely tied to large-scale system challenges, including child welfare, public health and public safety, sit outside the department, run by nonprofit groups that contract with the City.
What Dalton has demonstrated repeatedly is an ability to work across those silos, regardless of whether the systems are under her purview, using partnerships and shared data to tackle problems that no single agency can solve alone.
In a city as large and complex as New York, that kind of leadership is invaluable.
I consider this appointment a vote of confidence in evidence-based governance and in the idea that cities should reward leaders who test ideas, rigorously measure results, collaborate across systems and enable responsible access to integrated administrative data for external researchers so evidence can accumulate and lessons can travel across jurisdictions. That’s welcome coming from Mamdani, who some worry is going to be guided more by ideology than a desire to innovate and get results.
If the culture she helped build travels with Erin Dalton, all of New York City will benefit, not just New York City residents served under DSS. And importantly, the rest of the country will benefit too. In the nation’s largest city, results don’t remain local, they shape the field. And every other city loses the excuse that it can’t be done.
