Panel Paper: Medicaid Eligibility and Healthcare Utilization in Arizona: A Natural Experiment

Saturday, November 9, 2019
I.M Pei Tower: Majestic Level, Majestic Ballroom (Sheraton Denver Downtown)

*Names in bold indicate Presenter

Patrick Wightman, University of Arizona, Matthew J. Butler, Brigham Young University and William G. Johnson, Arizona State University


Given the disadvantages Medicaid beneficiaries typically face, the loss of coverage can have harmful consequences. Medicaid disenrollment has been found to be correlated with increases in emergency department use and health care costs. Untangling the threads of causality in these relationships is difficult. To address this concern we exploit a natural experiment arising from a policy shift that occurred in Arizona. In 2011, following the Great Recession, Arizona faced with the second worst budget shortfall of any state. In response, the legislature froze enrollment in its Proposition 204 program which enabled the state’s Medicaid agency to extend coverage to childless adults with incomes less than 100% of the FPL. By December 2013 enrollment through this program had fallen from its pre-freeze peak of approximately 230,000 to less than 70,000.

This project, made possible by funding from the Robert Wood Johnson Foundation, follows the health-care utilization and insurance status of those Arizona Medicaid beneficiaries who lost their coverage as a result of this policy. To do this, we leverage access to enrollment data from the Arizona Health Care Cost Containment System (AHCCCS, the state Medicaid program) matched to Hospital and Emergency Department discharge data from the Arizona Department of Health Services (ADHS). Records are matched using administrative sources and cover the period 2008 through 2016, through the re-institution of the Prop 204 program in 2014 made possible by the Affordable Care Act. The longitudinal structure of the data means that beneficiaries act as their own controls, mitigating the impact of unobserved confounding factors on the analysis. Moreover, the extended time frame allows us to examine any impact of pent-up demand following lifting of enrollment freeze.

Specifically, we investigate the impact of AZ’s Medicaid enrollment freeze on emergency department visits and inpatient stays, and the extent to which the loss of coverage causes erstwhile beneficiaries to substitute the former for the later. We also investigate the impact of renewed coverage on these behaviors.

These questions are especially timely given current legislative debates over Medicaid funding and eligibility criteria occurring at both the federal and state levels. While ACA-funded Medicaid expansion has made health insurance coverage available to millions of individuals, many for the first time, a number of states are considering conditioning availability on work and training activities, among other qualifications. This project will provide much-needed evidence to this debate.