Learning agendas are an up-and-coming tool to strengthen a culture of evidence-based decision-making within public agencies. While at the federal level they’re now required of many large departments, they’re just starting to be used by state and local agencies. Examples include the Colorado Department of Public Health and Environment and several agencies in North Carolina.

What is a learning agenda? It’s a document developed by agency staff or with the help of external research partners that identifies an agency’s priority research questions, such as: Is Program A effective? Which version of Program B is most cost-efficient? What can we learn from existing research about initiative C’s likely effectiveness? And what operational changes might make policy D more customer-friendly?

Identifying questions like these helps agencies and their research partners, such as universities, focus their evidence-building resources on the agency’s most important challenges. Sometimes learning agendas are a list of research questions, while others proceed one question at a time.

One might assume that the best way to start a learning agenda is just to ask agency leaders what types of research projects they think would be most useful. In practice, however, that approach can backfire. The term “research projects” can sound too slow and time-consuming — too much like an academic exercise, or more about theory than practice — to be useful for decision-makers. Also, many public officials don’t have experience thinking about formulating research projects.

A better approach, we’ve found, is to ask those leaders a simpler question: What keeps you up at night? In other words, what are the most pressing problems or challenges that the agency faces? Every leader will, most likely, have at least a few pressing issues to share, given the scope and complexity of the work that public agencies do.

It’s then up to the staff or external partners to develop a useful research agenda that could inform those challenges. Projects might use low-cost, rapid experimentation to test operational improvements, data analysis to provide relatively quick insights from existing data, program evaluations to learn what works, or a review of evidence from prior studies to draw on what’s already known. The exact projects — and the speed of completing them — will depend, in part, on what data and internal and external resources are available to carry out the research.

A recent example comes from North Carolina. The state’s Office of Strategic Partnerships (where one of us, Jenni Owen, is the director) met with leaders of the North Carolina Department of Public Safety to ask the “what keeps you up at night?” question. The discussion revealed three priority questions related to the COVID-19 pandemic.

  • How can the department and its stakeholders best provide access to health care and mental health resources for offenders re-entering the community during the pandemic?
  • What are best practices in prison supervision during the pandemic, including appropriate pro-social-behavior activities such as education, counseling, treatment and self-help?
  • What is the impact of employment barriers created by the pandemic on ex-offenders' successful transition from prison back to their communities?

With that information in hand, the Office of Strategic Partnerships identified research teams from three North Carolina universities that were interested in helping to tackle those questions. The teams reviewed existing research and analyzed public data. (They included graduate students, helping train the next generation of policy researchers.) Within six weeks, and at no cost to the state, the researchers delivered briefs with their initial findings. The department’s leaders were able to use the findings to inform their near-term decision-making while determining whether to request more in-depth analyses.

Examples like these highlight the potential impact of even brief, focused conversations with leaders about what keeps them up at night. Those conversations can inform the creation of learning agendas that, sometimes quite quickly, can produce data-driven insights to address key challenges.

When one set of questions is answered, new ones can be added, keeping up the cycle of learning. Moreover, those developing the learning agendas can start to seek broader input, including asking program managers, external implementation partners and other stakeholders what keeps them up at night. That way, a learning agenda can continue to identify important opportunities for better agency results — the kind of evidence-informed results that improve lives and provide better services.


Governing’s opinion columns reflect the views of their authors and not necessarily those of Governing’s editors or management.