Most access to justice innovations are chosen based on intuition, funded based on narrative, implemented without fidelity monitoring, and never rigorously evaluated. This session tackles the full evidence lifecycle — not by turning every justice professional into a researcher, but by providing practical tools for making evidence-informed decisions at every stage of the innovation cycle.
The session is structured in three acts. Act 1 highlights how researchers and court administrators can work together to develop and test reforms in the civil legal system. A case study of a partnership between Stanford Law School's Rhode Center on the Legal Profession and the Superior Court of Los Angeles County will demonstrate the benefits and challenges of designing and testing field experiments, policy reforms, and technological innovations that leverage court data to increase access to justice. Attendees will have the chance to apply what they learn to their local contexts, including an exploration of the possibilities and barriers to evidence-based reform in their own backyards.
Act 2 addresses why good ideas die in implementation. A case study in implementation science shows how one organization used validated assessment tools to match service intensity to need, structured staff engagement to shift organizational culture, and continuous quality improvement embedded into daily operations — not as an evaluation afterthought, but as how the work gets done. The methodology is universal: every court rolling out a new process, every legal aid organization launching a program, every self-help center redesigning its workflow faces the same question of how to ensure that an evidence-based model actually gets implemented with fidelity.
Act 3 shows what becomes possible when AI supports evidence work at scale. A live demonstration of
Redo.io, an AI-powered platform, reveals how technology can synthesize complex statistical analyses into coherent narratives that support California Racial Justice Act claims. The payoff is democratized access to sophisticated analyses of racial disparities in sentencing outcomes. Producing these analyses independently would be prohibitively costly and time-consuming for under-resourced public defender offices and other legal practitioners, who are often the ones who need them most. The demonstration also shows the validation work behind the platform: how the evaluation studies are designed, what we find, and the limitations of both the studies and the findings.
The session closes with a hands-on activity. Each participant picks one justice innovation they're currently working on and builds a mini evidence plan using a template that walks through three questions drawn from the three acts: Is this worth pursuing, and what does the data say? How will I ensure implementation fidelity once it launches? How will I measure whether it actually worked? Pairs review each other's plans, because the most useful feedback comes from someone who doesn't share your assumptions.
You'll leave with an evidence triage framework, an implementation fidelity checklist, a research design template, and a plan for one real initiative that you've already started pressure-testing against the evidence.
