Access Type

Open Access Dissertation

Date of Award

January 2019

Degree Type


Degree Name




First Advisor

Allen C. Goodman


We estimate the effects of having health insurance on the probability of facing trouble in accessing provider’s care and that of being refused as a new patient by the provider’s office in chapter 2 of this dissertation. The NHIS data for 2011-2015 are pooled. To fight the potential endogeneity problem arising from the selection bias associated with insurance status we use a 2SLS method with self-employment status of individual and family income as the instruments. The estimated results show that having health insurance reduces the probability of facing difficulty in getting provider’s appointment by 7.3 percentage points. On the other hand, it reduces the probability of being refused as a new patient by doctor’s office is 6.3 percentage points. The above results are found controlling for all predisposing, enabling, and need factors.

In the third chapter, we estimate the impact of the implementation of Medicaid expansion under the Affordable Care Act on provider’s practice in the ED of Iowa and Wisconsin hospitals. Difference-in-difference technique and 2013-2014 data from the State Emergency Department Databases are used to capture the causal effect of coverage expansion on ED treatment practice. Our results show that the number of procedures performed increased by 0.63 per visit and the number of diagnoses decreased by 0.49 in Iowa after the implementation of expansion coverage. We find the evidence of a parallel substitution towards higher complexity visits from an event study that compares the differences between the levels of ED visit complexities. Our results also show that the number of primary-care treatable and preventable ED visits decreases by 0.0067 and 0.0043, respectively. These results are suggestive of an appropriate shifting of patients from ED to other health care settings.

Included in

Economics Commons