Academic Commons

Theses Doctoral

Essays in Applied Microeconomics

Lee, Ajin

This dissertation consists of three essays in applied microeconomics. Each chapter covers a large category of public spending in the US: (1) health care; (2) social insurance; and (3) education. This dissertation aims to understand the determinants of efficient delivery of public programs, focusing on disadvantaged subpopulations.
The first chapter looks at the effectiveness of health care systems. Medicaid, the largest public health insurance program in the US, has transitioned from a fee-for-service system (FFS) primarily administered by the government to a managed care system (MMC) administered by private insurers over the last few decades. I examine how hospitals' responses to financial incentives under these two systems affect hospital costs and newborn health outcomes. I analyze the universe of inpatient discharge records across New York State from 1995-2013, totaling 4.5 million births. First, I exploit an arbitrary determinant of MMC enrollment: infants weighing less than 1,200 grams were excluded from MMC and were instead served through FFS. Using a regression discontinuity design, I find that newborns enrolled in MMC stayed fewer days in hospitals and thus had less expensive visits relative to newborns enrolled in FFS. The cost difference is driven by birth hospitals retaining more newborns enrolled in FFS while transferring away those enrolled in MMC. I find that MMC had limited impacts on newborn health, measured by in-hospital mortality and hospital readmission. Hospitals tended to transfer out MMC newborns only when a high-quality hospital was nearby, which resulted in these infants receiving uncompromised care. Second, I exploit county-level rollout of the MMC mandate to examine impacts on the full population of infants using a difference-in-difference design. I find that hospitals achieved a similar rate of cost savings as for infants over the 1,200-gram threshold, while length of stay, the probability of transfer, and mortality did not change following the mandate. This finding suggests that there are alternative, successful methods by which hospitals reduce costs under MMC, including for high-risk deliveries.
The second chapter argues that wealth uncertainty influences when couples choose to retire. Using data from the Health and Retirement Study, I show that wives delay retirement when their husbands retire following a job loss. This effect is stronger when husbands are the primary earners, and couples are relatively poorer. This provides evidence of intra-household insurance that mitigates the impact of an unexpected earnings shock. I find that wives tend to delay retirement only until they become eligible for Social Security. This suggests that Social Security benefits can relax households' budget constraints and allow wives to join their husbands in retirement.
The third chapter focuses on heterogeneity in grade retention decisions in New York City public schools. Performance on proficiency exams can be a key determinant of whether students are retained or "held back" in their grade. We find female students in New York City are 25% more likely to be retained in their grade due to exam failure than boys. Hispanic students are 60% more likely and Black students 120% more likely to be retained due to exam failure (relative to White students). Poverty and previous poor performance also increase the likelihood of retention, while being young for grade or short does not. We conclude that "patterned discretion" exists in how standardized test results are utilized.

Files

  • thumnail for Lee_columbia_0054D_13893.pdf Lee_columbia_0054D_13893.pdf application/pdf 750 KB Download File

More About This Work

Academic Units
Economics
Thesis Advisors
Almond, Douglas
Degree
Ph.D., Columbia University
Published Here
July 23, 2017
Academic Commons provides global access to research and scholarship produced at Columbia University, Barnard College, Teachers College, Union Theological Seminary and Jewish Theological Seminary. Academic Commons is managed by the Columbia University Libraries.