Karen Zhang

Economist

About me

I am a health economist studying how people, providers, and insurers respond to incentives – and what that means for cost, access, and quality.

I research health economics and teach practical tools to evaluate policy and markets

Day-to-day healthcare decisions rarely follow tidy theory. My work focuses on how choices are actually made – under time pressure, uncertainty, human error, and institutional constraints – and how policy or market shifts ripple through behavior into outcomes that matter.

As a dedicated educator, I bring that same focus on context and complexity into the classroom – treating empirical methods as tools for navigating imperfect information, institutional frictions, and practical trade-offs. Across all levels of teaching, my emphasis is practical judgment as much as technique: how to frame a question, pick a design that fits the setting, test whether assumptions are doing real work, and explain findings clearly enough that someone can act on them.

What I Study

My research sits at the corners of the healthcare triangle – payers, patients, and providers – highlighting the real-world responses that end up being more surprising (and more interesting) than the standard textbook story. Check out the following projects.

Payers/Insurers

Profit-maximizing insurers often respond to policy design in ways policymakers don’t anticipate – shifting pricing and participation as much as benefits. My dissertation examines the ACA exchange CSR subsidy shock and traces its unintended effects on premiums, competition, and welfare.

Patients

Lower prices and expanded access don’t simply increase utilization; they change which care people seek, when they seek it, and who becomes the marginal patient. A current project uses China’s centralized procurement reform to track how large device price cuts reshaped utilization and downstream care.

Providers

When staffing changes, care delivery adapts through triage, workarounds, and learning under pressure – often with measurable consequences for patients. Ongoing work isolates the effects of nurse turnover and team familiarity using quasi-experimental variation in staffing patterns.

Selected Research Projects

Dissertation

Cost-Sharing Reduction Subsidy: Who Benefits and How It Should be Financed, Evidence from the California Health Exchange

- Quantified the unintended consequences of a federal ACA marketplace subsidy defunding event on insurance coverage outcomes, plan-level competition, and total subsidy spending.
- Developed a structural model with endogenous insurer exit, allowing dynamic firm responses to profit conditions. Simulated counterfactual subsidy policies to assess impacts on market structure and participation.
- Computed welfare across alternative funding scenarios by integrating consumer surplus, public expenditure, and insurer profits, providing a comprehensive evaluation of redistribution and efficiency tradeoffs.

with Robert Huckman, Ingrid Nembhard, Amelia Bond & Steve Schawb

Staffing Mix, Turnover, and Productivity: Evidence from Military Nursing Wards

- Theorized and empirically validated team familiarity as a productive but episodic organizational asset that is accumulated through collaboration and depleted by turnover.
- Leveraged a Military Health System natural experiment to show that loss of team familiarity significantly increases patient length of stay, readmissions, and ICU admissions, with effects nearly doubling in complex care settings.
- Demonstrated that permanent staff contribute most to productivity, while temporary contractors contribute the least, providing causal evidence for the value of stable teams.

with Hanming Fang, Gordon Liu, Ruochen Sun & Huyang Zhang

Price, Access, and Quality under Centralized Procurement: Evidence from China's National Stent Procurement

-Analyzed the local impact of China’s 2021 national centralized volume-based procurement reform for coronary stents, which shifted nearly 90% of the local market to centralized purchasing.
- Applied a reduced-form design, starting with a Regression Discontinuity and extending to a Difference-in-Discontinuities framework, to estimate changes in prices, utilization, and provider behavior.
-Found large cost savings and expanded access, but also evidence of throughput-driven quality trade-offs, especially for complex patients, following the sharp price shock.