Return to the lead support bureau

Evaluating LEAD’s Impact

Quick Guides

Data-driven decision-making is more than just a catchphrase for LEAD. A strong and intentional evaluation plan, tracking progress towards clear goals, is a powerful tool that can advance consistency, sustainability, and impact. As public officials in King County, WA, have observed, compared to other diversion programs, LEAD stands out for its use of evaluation and data to assess impact and to identify opportunities for design modifications and improvement. The commitment to use data to measure impact and inform design changes is a core tenet of the LEAD framework.

Evidence shows that when it’s implemented with fidelity, LEAD works: A rigorous set of external evaluations of the flagship site in Seattle found that LEAD reduced rates of re-arrest by 58%, new felony charges by 39%, and prison admission by 87% while reducing systems costs and increasing rates of permanent housing by 89% and legitimate income by 33%. Those gains were experienced comparably across racial categories, indicating that LEAD was achieving the intended goal of reducing racial disparity among those subjected to punitive sanctions–not inadvertently exacerbating it, as is too often the case with well-intended reform initiatives.


Is anyone eligible for LEAD?

Many people who come into occasional contact with the criminal legal, social service, or behavioral health systems – via a single arrest, a temporary economic hurdle, or a difficult psychological period – can successfully find their way through these challenges without suffering severe and lasting consequences. For them, the established system of response and care may prove to be sufficiently accessible and manageable.

These are not the people LEAD is intended to serve.

Instead, LEAD is expressly designed to provide a new system of care for people whose complex, ongoing, unmet behavioral health needs result in disruptive or unlawful behavior. They may lack reliable shelter, income, food, health care, and positive social networks, and may find existing systems inaccessible, impossibly complicated, or insufficiently responsive.