This article is part of a series of interviews with recipients of Clinical Research Forum’s Top 10 Clinical Research Achievement Awards. This interview is with Eric T. Roberts, PhD, Associate Professor of Medicine, University of Pennsylvania. Dr Roberts is a health economist who studies the delivery and financing of care in U.S. public insurance programs and health care delivery systems. He received the 2026 Distinguished Clinical Research Achievement Award for “Loss of Subsidized Drug Coverage and Mortality among Medicare Beneficiaries.” The interview has been edited for length and clarity.
What early influences shaped your decision to pursue a career as a health economist?
From a young age, public policy was important to me. Both of my parents are artists, and when I was growing up, we didn’t have much money. For a while, my sister and I were on the Children’s Health Insurance Program (CHIP), and I was definitely aware that I was a beneficiary of a public program. I remember knowing my parents had concerns about what it cost for them to visit the doctor, but I went to the pediatrician’s office, my dad would say, “Oh, he has CHIP,” and that made things different for me. Experiences like that made me curious about how programs like these worked.
Was your focus always on health policy?
Actually, I was a history and economics double major at Johns Hopkins and initially, my focus was on the social determinants of economic outcomes. Going from neighborhood to neighborhood in Baltimore, you can’t help but be aware of the massive variations in economic prospects and, sadly, how strongly correlated those economic conditions are with racial segregation. One semester, I did an internship for the mayor’s office in Baltimore, and that got me working with local schools and health clinics. I also had friends who were public health majors, and through them I learned how to start thinking about these issues in a systematic way. So healthcare was always in my range of vision, but it wasn’t until college that I realized it could be an applied direction for my interest.
So, you were looking for a way to apply what you were learning in your economics classes?
Yes, although I wouldn’t say I’m enamored with economic theory. I like it as a tool, but I am much more interested in the application. My favorite economics classes in college were public finance, which is about how governments raise revenue, manage budgets, and make choices about what to fund, and econometrics, which teaches the statistical tools economists use for applied research.
It seems like the intersection of public finance and healthcare has become increasingly important
Yes, because the issues around how we can afford healthcare and who can’t afford healthcare have become ever more salient. The Affordable Care Act (ACA) was the culmination of the conversation that grew out of the Great Recession and rising uninsurance, and it tried to make broader coverage more economically sustainable. Ultimately, the ACA made important progress in addressing the coverage question, but it didn’t solve the issue of affordability – costs continue to go up. I always get questions about how we can make healthcare affordable, and the fact that there is no silver bullet answer is both a source of frustration and a source of fascination.
Is that what motivated the award-winning research?
This paper is the culmination of about eight years of work I’ve done on affordability of care and Medicare. The U.S.’s approach to healthcare assistance is a patchwork of programs that has developed incrementally, and a feature of this architecture is that there are a variety of different eligibility rules and arbitrary distinctions about who does and does not get help. This policy environment creates an enormous amount of administrative complexity that results in people either not qualifying for or periodically getting kicked out of these programs. Studying that revealed to me that not only do we have a cost problem in the US, but we also have an administrative burden problem. It’s not ill intended – it’s because there’s such a decentralized set of programs that serve different groups – but the result is that often the people who need the most help are subjected to the most complex architecture for getting that help. Sometimes these administrative restrictions are seen as benign ways of preserving program integrity, but they can also be blunt instruments, and I wanted to better understand the potential harm they cause.
How can you investigate that?
It is difficult to design studies that quantify the health effects of health coverage. You need very robust methods and the ability to compare large numbers of people to study outcomes like mortality. One way we can learn about the value of healthcare programs is by studying what people lose – for example, because they don’t qualify for these programs or because they are dropped from coverage for administrative reasons. For example, my study looked at what happens when low-income older adults lose Medicaid and with it, prescription drug subsidies. We identified dual-eligible Medicare–Medicaid beneficiaries, who automatically receive the Low-Income Subsidy (LIS), which reduces cost sharing in Medicare Part D. For these dual-eligible beneficiaries, being disenrolled from Medicaid often results in losing the LIS as well. Because of how these two programs interact, people dropped from Medicaid just a few months apart can vary by up to 17 months in when they subsequently lose the LIS. This created a natural experiment that allowed us to look at what happens when people lose the LIS early, solely because of the date they were initially disenrolled from Medicaid.
What did the results show?
Those losing LIS earlier were more likely to die. The increase in mortality was especially pronounced for people who had chronic conditions, like cardiovascular disease, chronic lung disease, and HIV – conditions where adherence to medications is important for maintaining health.
Why are these results important?
First, over 13 million low-income older adults depend on these drug subsidies to be able to pay for their medications, and this study showed the importance of those subsidies in helping them keep access to medications that save lives. Second, the study also showed the role of Medicaid in keeping people enrolled in the LIS and that when people lost Medicaid, they lost financial assistance for their drug costs.
What are you working on now?
I’d like to further investigate the tradeoffs people make when they’re suddenly faced with higher out-of-pocket costs for medications. Research has shown that when this happens, people become extra sensitive to even small changes in their finances, like how much money is in their bank account on a given day. For example, some work has shown that low-income older adults will delay filling medications until their Social Security paycheck arrives. I’m curious what other ways these financial constraints might affect health – such as cutting back of food or home heating, or delaying medical care. In addition to doing more quantitative work to flesh out the other health effects of coverage shocks, I’d like to use simulation modeling to project what could happen if we changed policies to make coverage easier to keep. Modeling out scenarios would allow policymakers to understand what could happen, and who would benefit, if we took deliberate steps to reduce coverage loss and the cascade of adverse events that stems from it.
What role does teamwork play in your research and how can that be supported?
Research like this is absolutely a team project, and teams thrive on continuity and the ability to have conversations and learn things together over an extended period – which requires an infrastructure for research. All the people who worked on this paper have federally funded grants and benefit from the continuity and infrastructure this provides. The career development award that I had from AHRQ (Agency for Healthcare Research and Quality) gave me the time and space to build these types of collaborations and start working with the other co-authors on this team, including José Figueroa and Rachel Werner. Incidentally, Lilly Estenson, another one of our co-authors, was on an AHRQ training grant while a graduate student at the University of Southern California, when she and I connected over our overlapping interests in Medicare affordability and coverage issues. This paper was possible because we were able to cross-collaborate, share ideas, and gain insight, and that happened because we had the research funding and infrastructure do so.
What advice do you have for people entering the field of health economics?
Today’s research environment is challenging, and I would recommend two things. First, you need to keep perspective, which often means taking time off to reset and recharge. This is such an intense, immersive job where you’re doing all sorts of very varied tasks, and while that’s very stimulating, it’s also very draining. For me, taking a step back means traveling with my wife, cooking, or swimming. The second bit of advice is to realize that we need to be more entrepreneurial in how we go after funding and more collaborative in how we make use of resources. There is a tendency for organizations to want their own lab or their own infrastructure – and that’s always going to exist to a certain extent – but in the face of competing priorities and difficulty getting funding, something has to give, and that thing is the desire to compete on things that are more shareable.
How can research teams adopt a more collaborative mindset?
One thing we can do is work more on connecting people to each other and opening up resources so that we can cooperate more creatively. We already do this to some extent, but we need to lean into it more. I prefer not to think of a Principal Investigator as the pinnacle of a research lab, but instead, as of a part of a dense, multilayered network of collaborators at many career stages, who are all figuring out how to share in incremental but cumulatively effective ways. Creating more functional multilateral relationships will allow us to still do important work despite other constraints.
Open access
