Dissertation title:  "Essays on Health Insurance: Public Opinion and Consumer Behavior"In this dissertation, I consider health insurance coverage in the United States from two perspectives. I open with a high-level analysis of trends in public opinion that underpin current political debates around health reform. My second two papers use administrative data from the Massachusetts Health Connector to characterize enrollment behavior in the nongroup health insurance market.

In Paper 1, I examine how preferences among three discrete approaches to reforming health insurance in the U.S. relate to political values and attitudes, as well as perceptions about and individual experiences with the health care system. Using people who would like to keep and improve upon the Affordable Care Act as a reference group, I find that individuals who prefer replacing the current insurance system with "Medicare for All" are generally less satisfied with the status quo in health care and their own experiences with the health care system. By contrast, people who would rather devolve health reform to the state level have views about the status quo and their own experiences that are more similar to the "Keep/Improve ACA" group, but sharply attitudes toward the role of government in health care.

The nongroup market for health insurance, which serves individuals who don’t have access to coverage through an employer or public programs, is known to experience high levels of churn. However, the reasons for disenrollment from this market are poorly understood. In Paper 2, I aim to clarify these dynamics using an administrative data set from the Massachusetts Health Connector. To obtain causal estimates, I leverage a natural experience for the 2017 plan year and use an event-study approach to study enrollment consequences of introducing a nominal (< $10) premium for plans that previously did not require a monthly contribution from enrollees. I show that nearly one in five enrollees who face a new, nominal premium are terminated for nonpayment by the end of the new plan year. As a first-order matter, enrollment inertia is an important determinant of these outcomes; all affected enrollees had the option to switch into a zero-premium plan offered by a different carrier at the start of the plan year. Comparing termination rates across plans that imposed different-sized premiums suggests that this disenrollment is predominantly driven by the administrative burden of initiating and sustaining premium payments, though there does appear to be a modest role for affordability.

Paper 3 uses the same data set to evaluate how take-up of individual market insurance during the COVID-19 pandemic compared to enrollment during prior years. I show that new enrollees who entered the Massachusetts individual market during the COVID-19 special enrollment period were younger, on average, and more likely to enroll in unsubsidized coverage than mid-year enrollees in 2017-2019. I also offer suggestive evidence that the state's special enrollment period—which relaxed administrative burdens typically required to enroll mid-year—may have improved take-up of coverage.