The Impact of State Medicaid Expansion Under the Affordable Care Act on Health Insurance Coverage at the County Level

December 6, 2016
 

Download the Brief

Endnotes are available in the pdf version. Download

Key Findings

Medicaid Icon
Counties in states that did not expand Medicaid compared to counties in states that did experienced significantly smaller increases in non-elderly adult health insurance coverage between 2013 and 2015, even after controlling for other county characteristics.
Icon of a clipboard
Counties in states that did not expand Medicaid compared to counties in states that did had larger shares of the vulnerable populations.
icon of three people
Within states that did not expand Medicaid, counties with larger shares of vulnerable residents experienced smaller improvements in health insurance coverage rates than did counties with smaller shares of vulnerable residents.

Summary

Counties and states with large shares of uninsured risk having to contend with a range of health and economic impacts, such as reduced workplace productivity, unsustainable demands on emergency departments, higher tax burdens resulting from uncompensated care costs, and deteriorating health care quality due to reductions in public spending.1

In 2013, before the implementation of major provisions of the Affordable Care Act, 41 million U.S. adults age 19–64 had no health insurance. Coverage varies considerably by geographic location. For instance, in 2013 county-level coverage rates ranged from a high of 96 percent in Norfolk County, Massachusetts to a low of 57 percent in Willacy County, Texas.2

The purpose of Medicaid expansion under the Affordable Care Act was to make health care more accessible to low-income populations. By early 2015, 28 states had expanded Medicaid eligibility (see Figure 1). The expansion by some states but not by others provides a unique opportunity to examine the impact of this new policy on changes in health insurance coverage. Moreover, as the newly elected Republican President and the Republican-controlled Congress consider the future of health care reform, understanding the efficacy of components of the Affordable Care Act, such as Medicaid expansion, will be essential for continuing efforts to increase coverage rates and subsequently minimize the associated consequences of low coverage rates. This research identifies differences in changes in insurance coverage rates for non-elderly adults (age 18–64) from 2013 to 2015 between counties in states that did and did not expand Medicaid. The analysis also identifies the county-level factors that contributed to these differences. The year 2013 is used as the starting point because Medicaid expansion did not begin until January 1, 2014.3

Figure 1. States Expanding Medicaid As of March 1, 2015Figure 2. Health Insurance Coverage, 2013
 

Health Insurance Coverage Continues to Vary Across the United States

The Northeast and upper Midwest have the highest rates of health insurance coverage for non-elderly adults, while the lowest rates are predominantly in the South. This geographic variation in coverage persisted in 2015 after the implementation of major components of the Affordable Care Act.

Insurance Coverage Increased in Nearly All Counties Between 2013 and 2015

Most counties (3,092) experienced increases in cover- age rates from 2013 to 2015 (see Figures 2 to 4), with the majority experiencing increases between 0.1 and 10 percent. Figure 4 illustrates the geographic variation in changes in coverage. Note that counties with declines or only small improvements in coverage include counties that began with high rates of coverage in 2013.

Figure 3. Health Insurance Coverage, 2015Figure 4. Percent Change in Coverage 2013-2015
 

Larger Improvements Occurred in Medicaid Expansion States

Counties located in states that expanded Medicaid experienced significantly larger improvements in adult health insurance coverage rates between 2013 and 2015, even after controlling for several other county characteristics. These findings complement previous research that has demonstrated the relationship between health insurance coverage rates and the expansion of public health programs like the State Children’s Health Insurance Program.4

Improvements in Coverage Varied by Metropolitan Status

Counties across all five metropolitan status categories experienced increases in insurance coverage rates from 2013 to 2015 (Figure 5). Coverage rates continued to be the highest in large metropolitan counties. While non- metropolitan counties in 2015 reached coverage rates equal to large metropolitan levels in 2013, they contin-ued to lag behind. These findings for non-elderly adults are similar to previous research findings showing that, in most regions of the country, health insurance cover- age rates for children are lowest in rural areas.5

Counties in States That Did Not Expand Medicaid Had Larger Shares of Vulnerable Populations

Compared to counties in states that expanded Medicaid, counties in states that did not expand it had larger shares of non-Hispanic blacks and adults with incomes below 138 percent of the federal poverty line and had fewer residents with at least a college degree (Figure 6). In other words, states that chose not to expand Medicaid had larger shares of the vulnerable populations who were meant to have benefited most from Medicaid expansion.

Figure 5. Mean County Health Insurance Coverage Rates by Metro Status, 2013 and 2015Figure 6. Average Demographic and Socioeconomic Composition of Counties By Expansion/Non-Expansion of Medicaid Coverage
 
Counties With Larger Shares of Vulnerable Populations Experienced Smaller Improvements

In states that did not expand Medicaid, counties with larger shares of vulnerable residents (that is, people in poverty, people with low educational attainment, or the foreign-born) experienced smaller improvements in health insurance coverage rates than did counties with smaller shares of vulnerable residents. However, coun- ties in states that expanded Medicaid were protected from these exacerbated disparities.

Conclusion

Although health insurance coverage rates increased in nearly all counties between 2013 and 2015, increases would have been larger if more states had expanded Medicaid. This is particularly true given that the states that did not expand Medicaid have comparatively larger shares of vulnerable residents at greatest risk of not having health insurance. Even though the Federal government initially covered 100 percent of the cost of expansion, with plans to phase down to 90 percent by 2020 and beyond, the long-term cost of expanded Medicaid in states with larger shares of vulnerable populations may have been a factor in the decisions by states not to expand coverage. As a result, counties with large shares of vulnerable populations in states that did not expand Medicaid experienced smaller improvements in insurance coverage. And while the most-rural counties experienced, on average, larger increases in coverage compared to large metropolitan counties, they still lagged behind in overall non-elderly adult coverage rates.

Policy Implications

States refusing to expand Medicaid will continue to see lagging county-level coverage rates.

Any proposed revisions to the ACA, and especially the curtailment of Medicaid, would reduce county-level insurance coverage rates and thus require counties to find new ways to deal with an increase in uninsured non-elderly adults.

For counties with large shares of vulnerable populations located in states that did not expand Medicaid, leaders interested in reducing the impact of lack of health insurance coverage should focus on increasing access to low-cost health care and preventive health.

Data and Methods

The outcome variable in these analyses is percent change in the county-level insurance rate for adults age 18-64 from 2013 to 2015 (N=3,141). The health insurance coverage estimates are from Enroll America. The primary independent variable of interest is state decision on Medicaid expansion. The author used multilevel regression analyses that control for the log of the total county population. Results reported here control for percent foreign-born, percent non-Hispanic black, percent of adults age 18–64 living at or below 138 percent of the federal poverty line, percent of adults age 25 and older with a four-year college degree or more, percent of individuals age 16 and older who are unemployed, and percentages of the labor force employed in (a) educational services, health care, and social assistance; (b) manufacturing; (c) local, state, and federal government; and (d) accommodations, food, and retail. Analyses also controlled for U.S. Census region, metropolitan status, and the county baseline insurance rate in 2013. All county measures come from the 2009–2013 American Community Survey five-year estimates,6 except for percent employed in accommodations, food, and retail, which come from the 2013 County Business Patterns;7 rural-urban continuum codes, which come from the USDA Economic Research Service (2013)8; and county baseline insurance rates (2013), which come from Enroll America.9 Rural Urban Continuum Codes were aggregated into five categories: large metro counties (population greater than 1 million), other metro counties (medium/large metro), nonmetro counties with an urban population of 20,000 or more (large nonmetro), nonmetro counties with an urban population of 2,500-19,999 (medium nonmetro), and counties with urban population less than 2,500 (rural).

Endnotes

1. Institute of Medicine of the National Academies, A Shared Destiny: Community Effects of Uninsurance (Washington, DC: National Academies Press, 2003), http://www.nap.edu/catalog/10602.html; Institute of Medicine of the National Academies, Hidden Costs, Value Lost: Uninsured in America (Washington, DC: National Academies Press, 2003).

2. Enroll America (2015), https://www.enrollamerica.org/research-maps/maps/changes-in-uninsured-ra....

3. Five states (California, Connecticut, Minnesota, New Jersey, and Washington) and the District of Columbia expanded Medicaid in 2010 or 2011 under a pilot program.

4. See Michael J. Staley, “Public Insurance Drove Overall Coverage Growth Among Children in 2012,” Issue Brief #73 (Durham, NH: Carsey Institute, University of New Hampshire, 2014), http://scholars.unh.edu/cgi/viewcontent.cgi?article=1215&context=carsey.

5. Marybeth J. Mattingly and Michelle L. Stransky, “Rural and Urban Children Have Lower Rates of Health Insurance Coverage and Are More Often Covered by Public Plans,” Issue Brief #8 (Durham, NH: Carsey Institute, University of New Hampshire, 2009), http://scholars.unh.edu/cgi/viewcontent.cgi?article=1081&context=carsey.

6. U.S. Census Bureau (2014), American Community Survey: 2013 Data Release (2014), http://www.census.gov/acs/www/data_documentation/2013_release/.

7. U.S. Census Bureau (2014), County Business Patterns: 2013 Data Release (2013), http://www.census.gov/programs-surveys/cbp/data.html.

8. United States Department of Agriculture, Economic Research Service Rural Urban Continuum Codes, http://www.ers.usda.gov/data-products/rural-urban-continuum-codes.aspx.

9. Enroll America (2015).

Download the Brief

Endnotes are available in the pdf version. Download

About the Author(s)

Danielle Rhubart
Danielle C. Rhubart is an instructor at the Wright State University and an affiliate to the University of Dayton. Read More...

Acknowledgements

The author is grateful for support from the Population Research Institute at Penn State, which is funded by an infrastructure grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R24-HD041025). The author also thanks Enroll America, for its insurance coverage estimates, her dissertation committee (Shannon Monnat, PhD; Leif Jensen, PhD; Diane McLaughlin, PhD; and Stephen Matthews, PhD), as well as Michael Ettlinger, Curt Grimm, Michele Dillon, Amy Sterndale, Beth Mattingly, Bianca Nicolosi, and Laurel Lloyd at the Carsey School of Public Policy and Anonymous Donors.