Essays on Health Economics

The first chapter examines whether Medicare Part D—the largest expansion of Medicare in recent decades—had a lasting impact on retirement behavior by providing subsidized
prescription drug coverage to older adults. We employ a difference-in-difference-in-differences
framework that compares changes in labor supply among individuals aged 65–68 versus 55–64
in a treatment group with employer-sponsored retiree health insurance (RHI) only until age
65, relative to a control group with lifetime RHI, before and after the 2006 implementation
of Part D. Using data from Health and Retirement Study (2000-2018), we find that Part
D reduces full-time employment by 6.52 percentage points among the treated older group,
with approximately 31% of the effect driven by transitions from full-time to part-time work
within the same job. The response is most pronounced among those with multiple chronic
conditions, women, whites, non-Hispanics, the unmarried, and individuals without a college
degree. Additional analysis shows that these effects are weakened after the Affordable Care
Act, particularly among low-educated individuals in states that expanded Medicaid. This
suggests that Medicaid expansion diminished the marginal insurance value of Medicare Part
D by offering alternative public coverage, thereby reducing its effect on retirement behavior.
The second chapter examines the effects of increased trade between Japan and the U.S.
on mortality rates in the U.S. using a shift-share instrumental variables approach. Overall,
we find that an increase in Japanese imports is associated with higher rates of drug-related
mortality and cardiovascular disease (CVD) mortality, and lower rates of mortality from
accidents. These effects exhibit significant racial disparities. Specifically, a $1,000 increase
in import competition is associated with a 13.5% increase in drug-related deaths per 100,000
Black individuals aged 20-64 years old compared to a 7.8% increase among white individuals.
Additionally, a $1,000 increase in import competition is associated with a 3.0% increase in
CVD deaths per 100,000 Black individuals aged 20-64 years old, while there is no statistically
significant effect among whites. Effects on mortality rates from accidents are driven by deaths
among whites. Our findings also indicate that the rise in drug-related and CVD mortality is
concentrated among males and in regions with relatively high shares of employment in the
automobile or computer industries.
The third chapter examines the impacts of Targeted Regulation of Abortion Providers
(TRAP laws) on the supply of maternal care providers by exploiting the staggered enactment
of TRAP laws across states. Our findings indicate a significant decrease in the number
of Obstetrician-Gynecologists (OB/GYNs) by 3.38 to 3.55 per 100,000 females aged 15-
44 following the enactment of TRAP laws. Furthermore, the enactment of TRAP laws
affects the composition of OB/GYNs. The decline in response to TRAP laws is particularly
pronounced among OB/GYNs under the age of 34 and those between the ages of 55 and 64.
Newly graduated OB/GYNs, especially those from lower-ranked medical schools, also reduce
their presence in states enacting TRAP laws. Although we do not find significant changes in
applicants to OB/GYN residency programs or medical schools in response to TRAP laws,
in-state applicants tend to shift their applications from private medical schools, which are
often higher-ranked, to public medical schools.
The fourth chapter evaluates the density of maternal health care providers across urban
and rural ZIP Code Tabulation Areas (ZCTAs) and across ZCTAs with relatively high Black,
Hispanic, and white populations. The data show striking urban-rural disparities in OB/GYN
density, as well as differences in OB/GYN density across high white vs. high Black and high
Hispanic ZCTAs. Including nurse practitioners, physician assistants, and the estimated
ten percent of family practice physicians who provide maternal health care, in addition to
OB/GYNs, increases maternity care provider density nationwide but exacerbates urban-rural
and racial/ethnic disparities.