Electronic Theses and Dissertations




Weiwei Chen



Document Type


Degree Name

Doctor of Philosophy


Business Administration



Committee Chair

Albert Okunade

Committee Member

Cyril Chang

Committee Member

Andrew Hussey

Committee Member

Ebenezer Olusegun George


This dissertation comprises four essays. The first two essay investigates the sensitivity of two largest components of health care expenditure — hospital care expenditure (HOCEXP) and physician and clinical services expenditure (DOCLNEXP) — to the changes in income and how much of the estimated sensitivity is due to purchasing more care versus purchasing better care. Although the two essays share the same decomposition model, the estimation is different in the second essay due to data limitations. Using 1999 - 2008 panel data of the 50 US states, we estimate and decompose the income elasticity of HOCEXP and DOCLNEXP into its quantity and quality components respectively. Our findings suggest that the both HOCEXP and DOCLEXP rises have more to do with quality than quantity change. The results mimic the literature indicating that both hospital care and physician and clinical services are normal goods and technical necessities at the state level. The third essay analyzes the effect of insurance coverage on the likelihood of an emergency department (ED) visit being non-urgent or primary-care-sensitive (PCS). We analyze the Tennessee Hospital Outpatient Discharge Data for 2008 and identify non-urgent and PCS ED visits following a widely used ED classification algorithm. Our results of a logit quasi-likelihood model show that noninsurance is associated with higher probability of non-urgent visits and PCS visits when compared to private insurance. The predicted effect of insurance coverage under PPACA depends on the mixed structure of insurance types. The fourth essay explores the determinants and effects of confidence on academic and labor market outcomes using a rich-informed nationwide survey of graduate Management Admission Test (GMAT) registrants. We discuss several ways to define and measure confidence. Our results suggest that many confidence measures differ by race, gender, observed ability and managerial experience. These confidence measures have some predictive power in eventual academic outcomes and more so for labor market outcomes.


Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to the local University of Memphis Electronic Theses & dissertation (ETD) Repository.