Second-year Master in Public Policy students presented their capstone projects at the Policy Analysis Exercise (PAE) Showcase on April 28. Students developed policy recommendations for real-world clients on topics ranging from responsible use of artificial intelligence to preventing homelessness in Bogotá.
Learn more about some of the students’ projects below.
- Madison Marsh MPP 2026: The Cost of Survival: Structural Inequities to Pancreatic Cancer Care
- Naveen Krishnan MPP 2026: Outpacing The Threat | Developing AI-Enabled Defenses to PRC Cyberattacks for Mobile Network Operators
- Dominga Fantoni MPP 2026: Solidarity Unemployment Insurance Benefits Impact on the Chilean Labor Market
- Keith Brockhurst MPP 2026, Bernadette Hicks MPP 2026, and Aureo Mesquita MPP 2026: From Permit to Project: Practical Reforms to Unlock Richmond’s Housing Pipeline
Madison Marsh MPP 2026
The Cost of Survival: Structural Inequities to Pancreatic Cancer Care
What sparked the idea for your topic? Why was this topic important to you?
I lost my mom to pancreatic cancer when I was 17; she was 41. In her memory, I’ve led a nonprofit dedicated to supporting hospitals, patient funds, and research. Her experience is the reason I came to HKS to learn how to better serve cancer patients through policy and philanthropy, and to work toward more equitable outcomes globally.
What policy question(s) were you trying to address?
How can we improve pancreatic cancer survival through policy interventions beyond research investment?
“The success of any policy depends on how well it is adapted to the communities it serves, which is an insight I will carry forward in future policy and advocacy work.”
What policy recommendations did you provide to your client?
I focused on improving survival by strengthening access to high-quality surgical care. Specifically, I recommended reforming the Centers for Medicare & Medicaid Services (CMS) to implement minimum surgical volume thresholds for reimbursement similar to policies used in organ transplantation. This would help ensure patients receive care that meets established quality standards and aligns with NCCN guidelines.
What did you learn from the PAE experience and how do you see yourself applying those lessons in future work?
I learned that data alone is not enough and community insight is essential. While my research included analysis of nearly 230,000 patients from the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) dataset, it was the interviews, site visits, and patient perspectives that brought the data to life and informed more grounded policy recommendations.
Working across both the U.S. and Germany also reinforced how critical implementation is. The success of any policy depends on how well it is adapted to the communities it serves, which is an insight I will carry forward in future policy and advocacy work.
Naveen Krishnan MPP 2026
Outpacing The Threat | Developing AI-Enabled Defenses to PRC Cyberattacks for Mobile Network Operators
What sparked the idea for your topic? Why was this topic important to you?
Before HKS, I worked in crisis management, focusing on global conflict and attacks. As a polyglot across several languages, that work took me to Taiwan where I studied U.S.-China relations and also grew increasingly familiar with cyberattacks. This was around the time I started to study these problems on the U.S. side as well, analyzing Salt Typhoon and the fallout of U.S. telecom companies getting hacked. I decided to focus on 'countering' these threats in graduate school with my PAE, researching artificial intelligence with Jake Sullivan and leveraging my experience as a U.S. Navy intelligence officer to focus on this idea.
What policy question(s) were you trying to address?
My PAE addressed how Mobile Network Operators (MNOs) can defend against state-sponsored cyberattacks as artificial intelligence collapses traditional security timelines. Specifically, I identified the optimal degree of Al-enabled autonomous defense and the necessary speed of intelligence sharing to protect critical national infrastructure.
“Despite entering this process with a background in diplomacy rather than as a ‘cyber expert,’ I discovered that even the most complex problems can be solved starting with one dataset at a time.”
What policy recommendations did you provide to your client?
My piece had to do with the work I published in War on the Rocks, which details how companies can automate defense systems and what the government can do to promote these changes in the private sector. The day after I published my work, Anthropic's new model, Mythos, was released, which further confirmed my theory of the need of automated defenses. I crafted a set of recommendations to be used for my client, which was a research lab that advises the private sector.
What did you learn from the PAE experience and how do you see yourself applying those lessons in future work?
Through this project, I learned to tackle massive national security challenges by breaking them down into digestible pieces. Despite entering this process with a background in diplomacy rather than as a "cyber expert," I discovered that even the most complex problems can be solved starting with one dataset at a time. This experience taught me how to use probabilistic modeling to turn abstract policy debates into concrete design choices. It also motivated me to better communicate my research publicly, resulting in my publication in War on the Rocks.
Dominga Fantoni MPP 2026
Solidarity Unemployment Insurance Benefits Impact on the Chilean Labor Market
What sparked the idea for your topic? Why was this topic important to you?
I’m interested in behavioral economics and labor markets, so I knew from the get-go I wanted to take advantage of the PAE to study something in this intersection.
While researching labor market policies in my home country, Chile, I found that the non-contributory component of the unemployment insurance system offered generous benefits but also created behavioral responses worth studying empirically. Furthermore, the policy's eligibility rules offered potential exogenous variation I could leverage for a credible causal identification strategy.
This topic matters to me because behavioral responses are often overlooked in policy evaluation. Understanding how individuals respond to incentives, not just whether a program reaches its intended beneficiaries, is essential for conducting informed cost-benefit analyses and maximizing policy impact. This is particularly important in the case of unemployment insurance, which affects people during one of the most financially vulnerable periods of their lives: involuntary unemployment.
What policy question(s) were you trying to address?
My goal was to estimate the impact of the solidarity unemployment insurance on unemployment length in Chile.
What policy recommendations did you provide to your client?
After finding no statistically significant evidence of distortions in recipients’ unemployment duration, I recommended policy changes that expand coverage and improve targeting. First, I suggested simplifying the application process and implementing awareness campaigns to increase take-up. Second, I recommended reducing the number of months someone must have worked to qualify. This would benefit vulnerable groups, who tend to have fewer months of formal employment and experience longer periods of unemployment.
“Going forward, I am confident I will put these learnings to work in my new role in economic consulting, where rigorous data work and applied empirical methods are central to the job.”
What did you learn from the PAE experience and how do you see yourself applying those lessons in future work?
The PAE strengthened my applied econometrics skills and improved how I organize data. The data cleaning and matching process was particularly instructive. I merged four administrative datasets, constructed employment spells at the worker-month level, and handled duplicates. This all showed me that building a clean analytical sample requires as much judgment as the econometric analysis itself.
I also gained a deeper understanding of how labor market context shapes policy impacts. Conversations with my advisors and my literature review helped me see how features like informality rates, take-up frictions, and benefit generosity can fundamentally change whether a policy generates the behavioral responses theory would predict.
Going forward, I am confident I will put these learnings to work in my new role in economic consulting, where rigorous data work and applied empirical methods are central to the job. In the medium term, I hope to return to researching labor markets and social insurance, building on the empirical toolkit I developed through this project.
Keith Brockhurst MPP 2026, Bernadette Hicks MPP 2026, and Aureo Mesquita MPP 2026
From Permit to Project: Practical Reforms to Unlock Richmond’s Housing Pipeline
What sparked the idea for your topic? Why was this topic important to you?
Our project was personal because we worked with Aureo's hometown of Richmond, California. Aureo spent the summer as an intern with the city’s Planning Division and noticed that while city staff were committed to expanding housing opportunities in Richmond, there were not many housing developments under construction compared to neighboring cities.
The team’s charge was to better understand the barriers to building new units specific to Richmond and research potential solutions to address the housing crisis within current budget and staff constraints. It was important for us to ground the analyzes in practical solutions that could be both impactful and feasible.
What policy question(s) were you trying to address?
Our project looked at the institutional, financial, and regulatory barriers to housing production in Richmond, and what policy tools the Planning Division could leverage to support the construction of new housing units.
We first considered progress made by the city on addressing regulatory barriers, such as building height and density, parking requirements, and other regulations that can often slow down or even prohibit housing development. We then analyzed the financial considerations developers face when building more housing in Richmond under current market conditions. Finally, we reviewed institutional barriers, such as staff capacity, length of application process, and the role of citizen boards.
“We witnessed the complexities of the housing development process from multiple perspectives, understood the practical constraints and opportunities of city government, and honed tangible skills by applying our classroom lessons to real policy problems.”
What policy recommendations did you provide to your client?
We provided our client with four primary recommendations that balanced both practical feasibility and impact. In order to encourage housing development, we recommended that the Richmond Planning Division:
- Create positions for two housing policy fellows
- Merge the Design Review Board with the Planning Commission into one unified development advisory board
- Implement impact fee deferrals on residential development projects
- Put forward a proposal to allocate a portion of the recent Chevron refinery settlement money to a revolving loan fund for remediating contaminated sites
Each of these recommendations was specifically tailored to address an aspect of the barriers to housing development we identified through our analysis.
What did you learn from the PAE experience and how do you see yourself applying those lessons in future work?
All three of us left this process with new insights and new skills that will help drive our future work. We witnessed the complexities of the housing development process from multiple perspectives, understood the practical constraints and opportunities of city government, and honed tangible skills by applying our classroom lessons to real policy problems. As each of us expects to work in housing policy, and likely local government, within the near future, we are confident that these lessons will make us more capable, more nuanced policymakers with an eye to what it takes to bring housing to life.
This process also afforded us the opportunity to dive deep into the ins and outs of the housing development process. We heard from city staff, housing developers, regional agencies, and neighboring cities, who all shared their perspectives on what works and what doesn’t. At the same time, we read countless articles and researched our own data, survey, spatial, and financial analyses. We ultimately came away with a much broader understanding of what it takes to bring housing to life in a community like Richmond. We know it’s not just making the financials work, but it’s also about cultivating the right relationships, reducing uncertainty, navigating very specific timeframes, and understanding the community’s visions and desires.
Each of us will bring this complex understanding to our own future work in housing policy.
Photography by Lydia Rosenberg. Inline photos courtesy of Naveen Krishnan and Bernadette Hicks.
