Affiliate Spotlight: Jacob Wallace on strengthening Medicaid through evidence use and generation

In this Affiliate Spotlight, Jacob Wallace, Faculty Co-Director of the SCALE Lab at the Yale School of Public Health, reflects on his experience researching Medicaid, the promise of test-and-learn approaches, and his new role as Scientific Advisor of J-PAL North America's forthcoming Medicaid Learning Lab.

What drew you to conduct research in health economics, particularly in the Medicaid space?

I found Medicaid before I found health economics. I grew up in Jakarta, Indonesia, as the son of a development economist, and initially expected to focus on development. After college, however, I had the opportunity to work on Medicaid policy at the New York State Department of Health. It became one of the formative experiences of my life—the issues were interesting, and the mission to build an effective safety net resonated deeply with me.

My interest in Medicaid ultimately led me to health economics. Medicaid programs operate under tight budget constraints, and economics provides a framework for rigorously thinking about how best to use available resources. I was also drawn to the idea that variation in state policies could be carefully studied to learn what does and doesn't work.

How has working directly with state health agencies shaped the research questions you choose to study?

My time in government and my work with governments have deeply shaped my research.

When I left New York Medicaid for graduate school, I had no intention of studying Medicaid managed care (MMC), the model through which states deliver services via contracted private health plans. But when I spoke with former colleagues at the state, they informed me that New York was moving toward "Care Management For All" and MMC was the chosen vehicle. They urged me to study market dynamics in MMC, and those conversations sparked research ideas that grew into my dissertation and remain central to my work today.

More recently, working closely with Connecticut’s Medicaid program has helped me understand how states are implementing H.R. 1, the 2025 legislation that adds work requirements to Medicaid, among other changes. This work has surfaced important areas for research and opportunities for new academic-public partnerships to meet this moment and advance our understanding of how to build a smarter safety net. 

What do you believe is the role research can play in better understanding what works across the Medicaid space?

Research is critical to better understanding what works in Medicaid. Policymakers want to understand whether their programs and policies, such as payment reform initiatives or care management programs,  are effective, but often lack the bandwidth or technical capacity to evaluate them rigorously.

When I worked for the New York State Department of Health, we implemented so many new policies each year that we never had time to look in the rearview mirror and evaluate if they were working. As a result, past policies were often judged as successes or failures largely on anecdote, and these narratives carried significant weight in discussions of new policies. Research allows states to interrogate these stories with data and evidence, offering a more objective view of policy effects. 

2025 brought many changes across the Medicaid space. In such a rapidly evolving policy environment, what questions do you think researchers and practitioners are paying closest attention to right now?

There is significant attention on the programs and policies most impacted by H.R. 1—such as work requirements in Medicaid—and rightly so. This moment presents an opportunity for academic-public partnerships to build a “test-and-learn” model: evaluating a program, tweaking and improving it based on the evaluation results, and re-evaluating to consistently learn what works. The model can then help states collectively figure out how best to support their residents in navigating these requirements. Randomized evaluations (e.g., A/B testing) can be a powerful approach for determining which messages and modalities most effectively help residents navigate work requirements. 

The pressure H.R. 1 has placed on social service programs has also led to a renewed interest in efficiency. Policymakers, asked to do more with less, are seeking ways to make their programs more efficient and streamline services. This has increased demand for evidence-based approaches to rein in spending on the program's fastest growing program components, such as long-term services and supports.

What do you see as the most significant challenges Medicaid agencies face when trying to adopt test-and-learn approaches?

The day-to-day realities inside state Medicaid departments create challenges for adopting “test-and-learn” approaches. First, policymakers often lack familiarity with randomization and may be uncertain about when randomized evaluations are ethical, legal, and feasible. This is an area where sustained, trust-based partnerships are critical.

Second, Medicaid agency staff often lack the bandwidth and expertise to conduct randomized evaluations. One way researchers can mitigate this is by embedding themselves within government so research staff can operationalize and analyze evaluations, either independently or in partnership with state staff.

Third, states generally lack governance frameworks for conducting rigorous evaluations, such as Institutional Review Boards or advisory groups. While evaluations are treated as quality-improvement activities rather than research, it is important to work with states to develop approaches to oversee evaluations from clinical, beneficiary, legal, and policy perspectives. 

Beyond operational processes, state officials may be unfamiliar with using evidence from randomized evaluations (or other studies) to shape Medicaid policy. In the face of bureaucratic pressure and politics, whether the evidence from test-and-learn campaigns receives a proverbial “seat at the table” in policy decision-making remains a pressing challenge.

You have recently taken on the role of Scientific Advisor for the upcoming Medicaid Learning Lab. From your perspective, what gap does this effort aim to fill across some of the challenges you’ve described in Medicaid?

The Learning Lab is designed to support Medicaid program administrators and to bridge the gap between the rigorous research methods used in the research community and the practical decisions and constraints administrators face. Policymakers are eager to incorporate evidence into their decision-making, but the available research does not always align with the specific, time-sensitive questions that must be considered in program design and implementation.

My hope is that the Learning Lab will foster new partnerships between researchers and policymakers and guide them toward co-created research that addresses today's policy needs or tackles the enduring questions facing the safety net. And it will accomplish this by building technical skills in rigorous evaluation and creating structured opportunities for researchers and state partners to come together to understand each other's priorities and resources. By deepening partnerships between researchers and policymakers, the Learning Lab will also facilitate the translation of existing evidence into policy and help identify key gaps for both groups.

For more information on the forthcoming Medicaid Learning Lab, please contact [email protected].