The Effects of Language and Geography-Defined Groups on Health Insurance Choice
Policy Design, Analysis, and Evaluation
The objective of this study is to measure how language and geography-defined groups influence participation in public health insurance programs. The theoretical model in this paper shows how better information on insurance states, gleaned through language group contacts in one’s local area, can help individuals decide whether or not to take up a public benefit or remain uninsured. This study focuses on Medicaid-eligible adults and Medicaid/CHIP-eligible children who speak a non-English language at home, and uses pooled cross-sections of the 2008-2009 American Community Survey (ACS). Adapting an empirical method developed by Bertrand, Luttmer, and Mullainathan (2000), I define the main variable of interest as the interaction between contact availability, the density of an individual’s language group in an individual’s local area, and group quality, the information and preferences related to Medicaid that an individual’s language group may possess, as measured by the language group’s Medicaid take-up rate. The empirical framework also uses language group and Public Use Microdata Area (PUMA) fixed effects to control for observable and unobservable differences across language groups and local areas. The main results and sensitivity analyses strongly suggest that language and geography groups have a statistically significant impact on an individual’s probability of taking-up Medicaid/CHIP: For a policy change that increases Medicaid use by 1 percentage point, the network for these language groups will increase the probability of taking-up Medicaid by 10 percentage points for adults and 7 percentage points for children. As eligibility expands under the Affordable Care Act and more people in a given language group enroll in Medicaid/CHIP, the multiplier effect could lead to higher overall program participation than might otherwise might be anticipated in a scenario without non-market interactions. These results can also help policymakers target outreach funds towards uninsured non-English speakers who are eligible for public benefits.