How has the Affordable Care Act (ACA) affected disparities in health insurance coverage among different racial and ethnic groups? We know that the ACA has substantially lowered uninsured rates nationwide, 1 but previous research has documented that these overall declines also led to reductions in racial and ethnic disparities in health coverage rates. 2
For this data brief, we analyzed findings from the U.S. Census Bureau’s nationally representative American Community Survey to report on the most recent trends in insurance coverage broken down by poverty level and by race (see How We Conducted This Study). We also compare the ACA’s effects on coverage in states that took up the Medicaid expansion with states that did not. 3 And we show changes in insurance coverage in the two largest expansion states, New York and California, and in the two largest nonexpansion states, Texas and Florida.
The proportion of nonelderly adults lacking health insurance fell from 20.5 percent in 2013 to 12.3 percent in 2017, a decline of 40 percent, while racial gaps decreased.
All U.S. racial and ethnic groups saw comparable, proportionate declines in uninsured rates (Exhibit 1). However, because uninsured rates started off much higher among Hispanic and black non-Hispanic adults than among white non-Hispanic adults, the coverage gap between blacks and whites declined from 11.0 percentage points in 2013 to 5.3 percentage points in 2017. Likewise, the coverage gap between Hispanics and non-Hispanic whites dropped from 25.4 points to 16.6 points.
Uninsured rates fell between 2013 and 2016 for all racial and ethnic groups and income categories, and then in 2017 inched upward for most groups. Gains in coverage for all groups were greatest between 2013 and 2015 and continued, though at a lower rate, in 2016. These gains ended in 2017. There were modest increases in uninsured rates among non-Hispanic whites and non-Hispanic blacks between 2016 and 2017.
The biggest absolute reductions in uninsured rates occurred among Hispanic, black, and lower-income, nonelderly adults in Medicaid expansion states (Exhibit 2). Because of this, while disparities in coverage shrank in both nonexpansion and expansion states, the reduction in disparities was greater in the latter states. From 2013 to 2017, the coverage gap between blacks and whites in expansion states had dropped from 9.8 percentage points to 3.2 percentage points, and the corresponding gap between blacks and whites in nonexpansion states declined from 11.4 points to 6.2 points.
Hispanic people had the highest initial uninsured rate and experienced the greatest gains (an overall decline of 15 percentage points in uninsured rates and a nine-point decline in the gap with whites). Black people also had higher initial uninsured rates than whites and experienced greater gains (a 12-point decline in uninsured rates compared with six points among whites). Hispanic noncitizens (such as green card holders) also made gains in their insurance coverage, although this group did not qualify for Medicaid or for subsidies.
The importance of the Medicaid expansion in reducing disparities can be seen by comparing the experience of advantaged and disadvantaged groups in expansion and nonexpansion states (Appendix 2). Before the ACA expansions, the fraction of black and Hispanic people who were uninsured in states that would subsequently expand Medicaid was over 50 percent higher than the share of white people who were uninsured in states that would not go on to expand Medicaid. By 2017, the sharp declines in uninsured rates for both Hispanic citizens and non-Hispanic blacks in expansion states brought these rates below the rates for non-Hispanic whites in nonexpansion states.
These gains were important in improving overall coverage among Hispanics, whose public insurance coverage rates increased by more than six percentage points from 2013 to 2017 (Exhibit 3). Private coverage grew most among black non-Hispanics and among Hispanics (Exhibit 4). At the beginning of this period, the private coverage rate for whites was 30 points higher compared to Hispanics. By the end of the period, that gap had narrowed to about 24 points.
The differences are particularly dramatic for the Hispanic population (Exhibit 5). While coverage rates were largely stable in expansion states between 2016 and 2017, trends began to reverse in nonexpansion states among non-Hispanic blacks and non-Hispanic whites (Exhibit 6).
The importance of Medicaid expansion in narrowing disparities can also be seen in a comparison of uninsured rates in the two largest expansion states, New York and California, with the two largest nonexpansion states, Texas and Florida (Appendix 2). For nearly all groups, uninsured rates were roughly twice as high in nonexpansion states in 2016. Moreover, uninsured rates for the lowest-income black, non-Hispanic, and Hispanic citizens in California and New York were lower than the overall uninsured rates for all non-Hispanic whites in Florida and Texas.
There are persistent disparities in health outcomes among U.S. racial and ethnic groups. These disparities echo differences in access to quality health services. 4 Lack of insurance coverage, in turn, limits access to such services. While eliminating disparities in insurance coverage alone will not eliminate racial disparities in health, it is a key first step.
In this study, we find evidence for a link between expansion in access to coverage and equity in receipt of that coverage. Gaps in insurance coverage among racial and ethnic groups narrowed extensively after implementation of the ACA coverage expansions, and especially between 2013 and 2016. These effects were greatest in states that expanded Medicaid. Our results suggest that expansions of Medicaid in additional states would likely reduce disparities further.
Data used for this analysis were drawn from the 2013–2017 years of the U.S. Census Bureau’s American Community Survey (ACS). Our sample included adults ages 19 to 64. We defined racial and ethnic categories as non-Hispanic white, non-Hispanic black, and Hispanic, who may be of any race. We also group people by their income relative to federal poverty guidelines. We use the ACS variable for citizenship because citizenship status affects people’s insurance coverage within racial and ethnic categories. We categorize those with both public and private insurance — about 4 percent of respondents — as having public insurance.
