At EPIC, MSCAPP Students Used Data to Surface Overlooked Policy Questions

For MSCAPP students Sinan Grehan and Wendy Wang, a summer at the Energy Policy Institute (EPIC) at the University of Chicago was an opportunity to do more than strengthen their technical skills. It was a chance to see how rigorous empirical research can sharpen public understanding of environmental problems, from wetlands regulation to the hidden costs of climate change.

The M.S. in Computational Analysis & Public Policy (MSCAPP), offered jointly by the Harris School of Public Policy and the Department of Computer Science at the University of Chicago, is designed to prepare students to work at the intersection of computation and policy. At EPIC, Grehan and Wang each brought that training to complex research projects that demanded statistical reasoning, data engineering, and careful policy interpretation.

Grehan entered the summer with a clear interest in environmental policy. Before MSCAPP, both his internship and thesis research had focused on related issues, including climate migration and FEMA aid. EPIC offered a chance to continue that work in a setting where policy questions and empirical methods were tightly linked.

Working under Professor Hajin Kim of the Law School, Grehan supported research on two topics: the effectiveness of voluntary corporate sustainability reporting and the impact of Sackett v. EPA on wetland development. His role included reviewing drafts, replicating econometric models, and helping investigate difficult data questions as they arose.

One of those questions became the highlight of the summer. Grehan was tasked with comparing two dredging and filling permit datasets from separate federal agencies, the U.S. Army Corps of Engineers and the Environmental Protection Agency (EPA). At first, the two datasets appeared to describe the same thing, but they did not line up. Untangling why required close examination of the data itself, along with deeper research into how each dataset had been constructed.

“This may not sound that interesting at first, but to me it was,” Grehan said. “I really enjoyed the sense of solving this mystery, almost like detective work.”

That experience captured something essential about policy research: the work is not only about running models, but also about understanding what data is actually measuring and whether it can support the claims researchers hope to make.

Wang’s summer project approached environmental policy from a different angle. Working with EPIC faculty member Eyal Frank, she contributed to a project studying whether warmer winters increase deer populations and whether that effect leads to higher rates of animal-vehicle collisions across the United States.

The question may sound unexpected, but its implications are significant. Animal-vehicle collisions cost billions of dollars and cause hundreds of deaths each year, yet they are rarely included in the broader conversation about the economic and human costs of climate change. The project aims to build evidence for that overlooked category of harm.

Wang was responsible for building the climate and population data pipelines, integrating them with wildlife and collision data, and producing the merged dataset used in the project’s regressions. She also created visualizations to summarize changes in winter severity over time and ran first-pass analyses linking those shifts to deer harvest and collision outcomes.

“The work from this summer got our core results running, and they point in the predicted direction,” Wang said. “Harsh winters are associated with measurably fewer deer harvested and fewer collisions the following year.”

One of the most meaningful parts of the summer, she said, was the challenge of validation. After extracting and aggregating 1.7 million rows of weather data, she needed to find a credible way to verify the results, even without a straightforward published benchmark. That process required her to think carefully about the structure of the underlying data and what could count as an independent check.

“I found the verification work surprisingly satisfying,” she said.

For both students, MSCAPP provided the foundation they needed to contribute quickly and meaningfully.

Grehan found himself relying heavily on the advanced statistics sequence, especially concepts such as Ordinary Least Squares (OLS), fixed effects, and difference-in-differences. Those tools helped him evaluate research design choices, replicate models from scratch, and provide feedback on econometric work already underway. He also saw how MSCAPP had trained him to think more carefully about method selection rather than simply defaulting to the first available approach.

Wang said nearly every part of the first-year MSCAPP curriculum surfaced in her EPIC work. Debugging and testing skills helped her reason through large, messy datasets. Database training proved essential when joining multiple long-run datasets by county and year, especially given shifting county boundaries and changing retrieval methods over time. Statistics core informed how she thought about identification and how to separate meaningful variation from noise.

“This summer was the first time I did that outside the classroom,” Wang said, referring to building a project end-to-end. “I wouldn’t have been able to do any of this work a year ago, before CAPP.”

Grehan also came away with a stronger sense of how rapidly policy research methods are evolving. “I didn’t realize just how far econometrics had come for policy research, particularly with quasi-experimental research methods,” he said. “There’s a lot of relatively recent methodological developments that have enhanced their accuracy.”

That realization points to one of the clearest throughlines in both students’ summers: today’s policy research depends not just on technical fluency, but on judgment. In Grehan’s case, that meant asking whether a dataset or strategy was truly the right fit for the question. In Wang’s, it meant recognizing that even in an era of AI-generated outputs, the harder and more valuable task is still deciding whether those outputs are valid, interpretable, and useful.

Taken together, their experiences at EPIC reflect what MSCAPP is designed to do: prepare students to move between computation and policy with rigor, flexibility, and purpose. Whether tracing the effects of wetlands regulation or uncovering an overlooked climate cost, Grehan and Wang spent the summer showing how careful research can make public problems more visible and more understandable.

The experience also helped clarify what comes next. Grehan said the summer strengthened his interest in environmental research, consulting, and program evaluation. Wang, who had not expected to spend the summer in academic research, found that the work deepened the kind of rigor she had hoped to build when she entered MSCAPP.

For both students, EPIC offered more than a summer internship. It offered a view of what it looks like to do research that is technically demanding, policy-relevant, and grounded in questions that matter.