The Effect of ACA Medicaid Expansions for Adults on Insurance Coverage for Low-Income Children

A key goal of the Affordable Care Act (ACA) is to provide affordable coverage for health insurance for all Americans.  Among the programs implemented to achieve this outcome, the ACA provides an option for states to expand Medicaid coverage to low-income adults in families with incomes below 138% of the federal poverty level (FPL).  Alternatively, for children in these families, the early years of the ACA presented very few changes in eligibility.  The ACA requires states to maintain Medicaid income eligibility thresholds at pre-ACA levels through 2019 and includes a requirement to extend Medicaid eligibility to children in the 17 states that had covered children between 100-138% FPL using separate state CHIP in the pre-ACA period.  Despite the limited changes seen in Medicaid policies for children, there remains a potential for the ACA to have an impact on the probability a Medicaid-eligible child enrolls in public coverage.  Previous research on Medicaid-expansions to low-income parents showed that Medicaid-eligible children were more likely to enroll in public coverage when a parent was also eligible for Medicaid.  The literature also shows that Medicaid-eligible children were more likely to enroll in public coverage with the implementation of the CHIP – where it is believed that the outreach associated with CHIP was successful at increasing take-up among both CHIP-eligible and Medicaid-eligible children.  Known as “welcome mat” effects, these policies targeted to parents or other low-income families had spillover effects onto Medicaid-eligible children who otherwise had seen no change in their own eligibility status.      

This paper investigates the potential for “welcome mat” effects under the ACA.  In particular, we observe publicly eligible children and measure the impact of Medicaid expansions for adults on the probability that a child enrolls in public coverage.  Using data from 2013 and 2014, we take advantage of the natural experiment created by the Supreme Court decision that allowed states to choose whether or not to expand Medicaid to adults.  We use a difference-in-differences model to compare changes in eligible children’s coverage between 2013 and 2014 in Expansion-states vs. Nonexpansion-states.

We use the data from the American Community Survey (ACS) conducted by the U.S. Census Bureau.  The ACS is currently the largest household survey in the U.S., and draws a sample of three million U.S. addresses per year. In 2014, the ACS public use sample included 708,121 children. The ACS contains rich information on demographic, socioeconomic, geographic, and health characteristics, including income, citizenship, disability, and health insurance coverage. The information enables us to determine the eligibility of Medicaid (and CHIP) and the types of insurance coverage held for both children and their parents.

Preliminary results suggest the presence of welcome mat effects of the ACA.  Sample means for children in families with income below 138% FPL using ACS data show the increase in the probability of public coverage for all children in 2014 compared to 2013, and that this increase was twice as large in states that expanded coverage to adults than in states that did not.