The financing of long-term care services and supports (LTSS) relies heavily on self-insurance in the form of housing or financial wealth. Exploiting both local market variation in housing prices and individual-level variation in stock market wealth from 1996 to 2016, we show that exogenous wealth shocks significantly reduce the probability of LTCI coverage, without altering Medicaid eligibility among people with housing and financial assets. The effect of shocks to liquid wealth strongly dominates the effect of housing wealth changes. A $100 K increase in housing (financial) wealth reduces the likelihood of LTCI coverage by 1.24 (3.22) percentage points.

Expanding health insurance can either increase or decrease pregnancies and births, depending on how it affects access to contraception and the financial costs of childbearing. I study the Affordable Care Act's Medicaid expansion using difference-in-differences models and state-level data on contraceptive provision, births, and abortions, synthesizing analysis of these outcomes to distinguish the role of subsidized contraception. The expansion led to a substantial increase in contraceptive provision, particularly for short-acting hormonal methods. Teen birth rates fell significantly by 5%, a reduction that owes to increased parental eligibility, while effects on overall birth rates are not distinguishable from zero. Estimated effects on abortion are imprecise but suggestive of reductions for teens. Simulations based on contraceptive provision imply that subsidized contraception led to reductions in the pregnancy and birth rate of approximately 1% to 2%. These findings indicate that Medicaid's contraceptive subsidies modestly reduced pregnancies, abortions, and births, but also suggest that financial barriers to reproductive health care are not a primary driver of fertility behavior overall.

The overutilization of costly hospital and emergency care and the underutilization of primary care is considered a key driver of wasteful health spending. We study how a shift toward community-based primary care reshapes care patterns across levels of the health system in El Salvador. Using data on outpatient consultations and hospitalizations across 254 municipalities, we exploit the staggered introduction of community health teams (CHTs) between 2010 and 2013 using an event-study design. We find that CHTs increased preventive care, reduced curative consultations for conditions amenable to effective primary care, and lowered preventable hospitalizations. We also document improved primary-care coverage for previously unattended chronic conditions. Together, these results suggest that strengthening primary care through CHTs can meaningfully improve the organization of care and bolster health system performance.

We propose a theoretical model to study individual lifestyle choices related to calorie intake and physical activity, depending on personal fitness and body weight. The model builds on the rational eating literature and can generate a variety of behaviors that are consistent with the empirical evidence. In particular, we show that engaging in periods of a sedentary lifestyle can be a rational, utility-maximizing decision-a finding that is not present in the existing literature but is empirically widespread. Additionally, we show the possible existence of multiple equilibria and multiple indifferent lifestyles. The former justifies policy interventions to help individuals exit a self-reinforcing, but unhealthy equilibrium; the latter provides a theoretical basis for remediation plans that compensate for earlier unhealthy behaviors.

A large body of literature demonstrates that exposure to major adverse events such as natural disasters affects physical and mental health. Less is known about health consequences of long-term exposure to smaller, recurring shocks such as mining-induced earthquakes. Leveraging data from the Dutch Lifelines Cohort Study and Biobank and the Royal Netherlands Meteorological Institute, we examine mental health effects of frequent earthquakes generated by the extraction of natural gas, which was a major source of economic revenue for the Netherlands. Long-term exposure is captured by the accumulated peak ground acceleration. We employ individual-level fixed effects models to deal with selective exposure. We find that exposure increases depression and anxiety symptoms. Our results are robust to selective migration and to varying the exposure indicator. The results support a reassessment of the societal costs of the mining of natural gas.