This dissertation explores different topics on health economics. The first chapter quantifies the supply response of primary care physicians to a reimbursement bonus, the Physician Scarcity area bonus, introduced by Medicare between 2005 and 2008. The bonus represented a 5\% increase in reimbursement and it was paid to healthcare providers working in underserved areas of the U.S. I test whether this incentive attracted more physicians to these areas, if it affected the number of patients served, and if it changed the quantity of care provided by physicians. This is a relevant question as the country faces a shortage of primary care physicians that is expected to increase in the future. Using Medicare claims data between 2000 and 2010 combined with a difference-in-difference strategy, I find that the bonus had either a weak or undetectable impact on these outcomes. Descriptive studies and surveys suggest that the results could be explained by lack of information about the bonus and its small relative size. In the second chapter I explore health effects of the Flint water crisis on the elderly population. The water crisis of 2014-2015 exposed the residents of the city to high levels of lead and other contaminants. Much of the existing evidence of lead exposure comes from studies focused on children, as brains during childhood are more vulnerable than during adulthood. In April 2014, the city of Flint changed its water source from Lake Huron to Flint River, exposing its residents to high levels of bacterial and chemical contaminants, including lead. For approximately a year and a half, local authorities reassured residents that the water was safe for human consumption. Finally, in December 2015 the Mayor of Flint declared state of emergency. Using the universe of Medicare beneficiaries living in Michigan between 2012 and 2016, and the CDC's National Detailed Mortality Files between 2010 and 2017, combined with a difference-in-difference strategy and synthetic controls methodology, I estimate the effect of the change in the water supply on hospitalization and mortality rates. My results suggest a weak increase in all-cause and kidney disease hospitalization rates in Flint after the start of the crisis, and an increase in all-cause mortality rates. However, these results are usually smaller than what has been suggested by the epidemiological literature and the graphical evidence does not strongly support them.
