Improving Community Outcomes

What does it take to drive measurable, lasting change in a community? In this urgent and data-rich panel, social scientists and policymakers explored what works—and what doesn’t—in the fight to prevent homelessness and improve public outcomes. Hosted as part of LEO’s Evidence Matters series, the panel focused on King County’s Youth and Family Homelessness Prevention Initiative (YFHPI)—a bold attempt to move from emergency response to early intervention in the fight against homelessness.

Dr. David Phillips of LEO detailed the randomized controlled trial that evaluated the YFHPI, describing it as one of the strongest empirical cases to date for homelessness prevention. The results were striking: a 40% reduction in the likelihood that program participants entered homelessness, relative to the control group. This wasn’t an estimate or projection—it was a measured effect, born out of rigorous design and executed in real-world conditions.

But the session made clear: data alone is not enough. Carrie Cihak emphasized that public institutions must become “learning organizations”—agile, transparent, and deeply engaged with the people they serve. For her, evidence is a moral tool as much as a methodological one. Programs must not only work—they must work for the right reasons and in the right ways.

Kimberly Dodds and Tanya Robertson brought this commitment to life by grounding the conversation in the experience of families served. They described how flexible, low-barrier supports—often as simple as rental assistance or help navigating a school enrollment—can radically shift a family’s trajectory. These moments of upstream care are not just interventions, they argued, but acts of dignity and repair.

Vincent Quan of J-PAL placed this work in a broader national and international context. He noted that RCTs, while often difficult to implement, offer clarity in a field crowded by good intentions. The challenge, he said, is not simply to prove that something works—but to scale what works without losing the human relationships and localized wisdom that made it effective in the first place.

The conversation ended not with metrics, but with a call to imagination: What would it mean for our systems to be designed around prevention, rather than crisis? Around human flourishing, rather than institutional convenience? And what role does each sector—academic, nonprofit, public, and philanthropic—play in making that possible?

LEO’s Evidence Matters series makes one thing clear: poverty is not inevitable. But its solutions require courage, rigor, and above all, collaboration.