The Affordable Care Act Reduced Socioeconomic Disparities In Health Care Access

Access to health care among nonelderly Americans is strongly associated with socioeconomic characteristics, including income, education, employment, and wealth. ^1–5 Compared to Americans who are better off, those in lower socioeconomic strata are less likely to be insured, ^6,7 are more likely to avoid medical care due to cost ⁸ and to enter hospitals through emergency departments, ⁹ and have twice as many avoidable hospitalizations. ¹⁰ The poor use less health care in spite of having greater medical need. ¹¹ These health care access gaps are compounded by—and may contribute to—the large and widening socioeconomic disparities in health and longevity in the United States. ^6,12–14

The Affordable Care Act (ACA) was designed to improve access to health care by expanding insurance coverage. Although some aspects of the ACA applied to people of all socioeconomic strata—such as eliminating exclusions due to preexisting conditions—key features of the law sought to increase coverage among lower-income people specifically. These features included federal subsidies to expand eligibility for Medicaid to all Americans with incomes of up to 138 percent of the federal poverty level ¹⁵ and large premium subsidies for people with incomes of 100–400 percent of poverty who purchase insurance on the newly created exchanges. In January 2014 twenty-four states and the District of Columbia expanded Medicaid, and residents of all states gained access to subsidized premiums. Twenty-six states chose not to expand Medicaid at the time, though nonpoor residents of these states gained access to subsidized coverage on the exchanges. President Donald Trump and the Republicans in Congress have proposed repealing the ACA and eliminating many of these subsidies to lower-income people.

In this article we assess the extent to which the ACA—and its Medicaid expansion, in particular—reduced socioeconomic gaps in access to health care. Previous studies on the effects of Medicaid expansion suggest that health coverage increased, particularly for members of racial and ethnic minority groups, ^16,17 the poor, ^18–22 and younger adults, ²³ with gains concentrated in Medicaid expansion states. ²⁴ Furthermore, there is some evidence of increased use of preventive and primary care services in expansion states ^23,25,26 and a higher proportion of citizens reporting excellent health. ²¹ However, many existing studies have relied on surveys with very low (5–10 percent) response rates, used data from just a few states, assessed just the first year of full ACA implementation, or have not accounted for preexisting trajectories in outcomes. The effect of the ACA on socioeconomic disparities in access has not previously been reported.

Using nationally representative data for 2011–15 from the Behavioral Risk Factor Surveillance System (BRFSS), we assessed changes in health insurance coverage and access associated with the ACA for people in different socioeconomic strata, comparing changes between Medicaid expansion and nonexpansion states. We quantified changes in socioeconomic access gaps, defined as differences in access between low and high socioeconomic groups, in the two groups of states.

Study Data And Methods

Data

Data were extracted for all nonelderly adults (people ages 18–64) who responded to the 2011–15 BRFSS. For a description of the data, see the online Appendix. ²⁷

Several states (Arizona, Delaware, Hawaii, New York, and Vermont) and the District of Columbia provided health coverage to households with incomes at or above 100 percent of poverty before the ACA’s Medicaid expansion and were excluded from this analysis. ²⁸ We also excluded Massachusetts and Maryland because they had statewide programs covering adults who had no dependent children and whose household incomes were up to 150 percent ²⁹ and 116 percent ^30,31 of poverty, respectively. We did not exclude California because its pre-2014 Medicaid expansion was not statewide and did not always cover people with household incomes of at least 100 percent of poverty. Our final data set contained a total of 1,089,940 respondents from the remaining forty-three states. Summary statistics are presented in Appendix Exhibit S1, and a map of states by expansion status is presented in Appendix Exhibit S12. ²⁷

Measures

We assessed changes in three measures of health care access. Insurance coverage was measured by asking, “Do you have any kind of health care coverage, including health insurance, prepaid plans such as HMOs [health maintenance organizations], or government plans such as Medicare, or Indian Health Service?” Whether or not respondents had a primary care provider was measured by asking, “Do you have one person you think of as your personal doctor or health care provider?” Lastly, whether or not a respondent avoided care due to cost was measured by asking, “Was there a time in the past 12 months when you needed to see a doctor but could not because of cost?” These three measures have been found to have high levels of validity and test-retest reliability. ³²

Analyses were stratified by respondents’ socioeconomic characteristics: self-reported household income, educational attainment, employment status, and home ownership status. The BRFSS reports annual household income in eight categories: less than $10,000, $10,000 to less than $15,000, $15,000 to less than $20,000, $20,000 to less than $25,000, $25,000 to less than $35,000, $35,000 to less than $50,000, $50,000 to less than $75,000, and more than $75,000. Because of small sample sizes at lower incomes, we also stratified using a binary indicator for household poverty (household income of less than $25,000 per year), which allowed us to identify households most likely to have incomes below the federal poverty level (the poverty level for a family of four in 2014 was $23,850). In describing income-related access gaps, we compared people in households with higher incomes (more than $75,000) to those in households with lower incomes (less than $25,000). These categories each represented about 30 percent of the respondents.

Education was treated as binary characteristic: whether or not the respondent had graduated from college. Employment status was defined as employed, unemployed, or not in the labor force. Home ownership was a binary characteristic: whether the respondent’s household owned or rented its home.

To be considered an expansion state in this analysis, a state had to have implemented the ACA Medicaid expansion by mid-2015 (Pennsylvania and Indiana were considered expansion states; Alaska was not). Of the forty-three states included in the analysis, twenty-one were categorized as expansion states. In sensitivity analyses, we excluded all states that had expanded Medicaid after January 1, 2014: Michigan (which expanded on April 1, 2014), New Hampshire (August 15, 2014), Pennsylvania (January 1, 2015), Indiana (February 1, 2015), Alaska (September 1, 2015), and Montana (January 1, 2016).

Analytic Approach

Our analysis proceeded in four steps. First, we assessed how each of our outcomes (having insurance coverage, having a primary care provider, and avoiding care due to cost) varied with socioeconomic characteristics, thus illuminating disparities in access. We assessed these relationships in 2013 (before implementation of Medicaid expansion and the new health insurance exchanges) and 2015 (up to two years post-implementation), stratifying by whether each state had expanded Medicaid.

Second, we estimated pre/post “first differences” regression models to assess temporal changes in health care access associated with the ACA rollout. We estimated both crude changes (from 2013 to 2015) and adjusted changes (controlling for state-level time trends for the period 2011–15 and the following respondent characteristics: race/ethnicity, sex, age category, pregnancy status, veteran status, education level, home ownership status, household size, household income, presence of children in the household, and state). Models were stratified by socioeconomic characteristics and by residence in an expansion versus a nonexpansion state.

Third, to identify changes in outcomes associated with Medicaid expansion, we estimated difference-in-differences models, adjusting for national time-varying factors—including the rollout of other aspects of the ACA that were implemented in all states. We estimated both crude and adjusted difference-in-differences models, controlling for state-level trends for the period 2011–15 and the covariates listed above. Crude models included data just for 2013–15; adjusted models used data for 2011–15 to better capture pre-reform trends. The difference-in-differences models were stratified by socioeconomic characteristics. All regression models were estimated as linear probability models with BRFSS sampling weights ³³ and standard errors clustered at the state level to account for intrastate correlation. ³⁴ Regression equations are presented in the Appendix. ²⁷

Fourth, we assessed changes in health care access gaps between 2013 and 2015, defined as absolute and relative changes over time in the percentage-point difference in access between people in high and low socioeconomic strata for each socioeconomic characteristic. Our analysis of access gaps was stratified by whether the state expanded Medicaid. Absolute changes in access gaps were assessed in regression models, interacting the socioeconomic strata with the post-reform indicator. All analyses were conducted using R, version 3.24.

Limitations

Our study had several limitations. First, as with all nonexperimental studies, certain assumptions are required to interpret the estimates as causal. Our adjusted first-differences models could be interpreted this way if all secular changes between 2013 and 2015 were attributable to the ACA, after adjustment for linear time trends and changes in observed covariates. Our difference-in-differences models relied on the assumption that expansion states would have experienced changes similar to those in nonexpansion states had they not expanded Medicaid, after adjustment for state trends and covariates. In interpreting our difference-in-differences models, we note that if Medicaid expansion states differentially implemented other aspects of the ACA—such as more or less advertising and outreach ³⁵ —then our effect estimates could reflect these differences in addition to the direct effect of Medicaid itself.

A second limitation is that survey nonresponse could also be a source of bias. The BRFSS response rates are about 40–50 percent, which is high for telephone surveys but still indicates substantial nonparticipation. ³⁶ Although responses were reweighted to reflect state demographics, the data may be nonrepresentative in other ways. We adjusted for observed characteristics in our models to reduce the influence of variation in survey participation.

Third, the analysis included data from only the first two years of the ACA Medicaid expansion and exchanges. More distal outcomes, including health outcomes, might need more time to respond to this policy intervention. ^17,37–41

Finally, the persistence of the observed changes is uncertain, given the changing policy environment.

Study Results

The study sample was weighted to reflect the noninstitutionalized US resident population ages 18–64 years. Compared to Medicaid nonexpansion states, expansion states had smaller proportions of black residents and a somewhat higher average household income, but similar levels of employment and homeownership (for sample characteristics, see Appendix Exhibit S1). ²⁷

In 2013 there was a steep gradient in coverage across income groups: Over 90 percent of Americans in households with annual incomes of more than $75,000 were insured ( Exhibit 1 ), compared to only about 60 percent of Americans in households with annual incomes of less than $25,000 per year (63.2 percent in expansion states and 55.0 percent in nonexpansion states) (for more detailed results on insurance coverage by income group, see Appendix Exhibits S2 and S3). ²⁷ Steep income gradients were also observed in 2013 for access to a primary care doctor and avoiding care due to cost (for detailed results, see Appendix Exhibits S4–S7). ²⁷ Large pre-reform access gaps were also observed between education, employment, and home ownership strata (for results, see Appendix Exhibits S2–S7). ²⁷ By 2015 the income-access gradient had flattened substantially in Medicaid expansion states, with smaller changes observed in nonexpansion states ( Exhibit 1 ).

SOURCE Authors’ analysis of data for 2013 and 2015 from the Behavioral Risk Factor Surveillance System (BRFSS). NOTES The exhibit displays the percentage of noninstitutionalized US adults ages 18–64 who reported that they had insurance coverage, by BRFSS household income category. As explained in more detail in the text, to be considered an expansion state, a state must have expanded eligibility for Medicaid through the Affordable Care Act by mid-2015.

Changes in access from 2013 to 2015 differed by household income category. In Medicaid expansion states, there were large increases in insurance coverage for the poor under the ACA, but little change at higher incomes ( Exhibit 2 ). Gains in access to a primary care provider and reductions in avoiding care due to cost were also strongly concentrated among the poor and were about half the size of the gains in insurance coverage. (Changes for nonexpansion states are shown in Appendix Exhibits S2–S7.) ²⁷

SOURCE Authors’ analysis of data for 2013 and 2015 from the Behavioral Risk Factor Surveillance System (BRFSS). NOTES The exhibit displays changes in the percentage of noninstitutionalized US adults ages 18–64 who reported that they had insurance coverage, had a primary care provider, and did not avoid care due to cost, by BRFSS household income category. As explained in more detail in the text, to be considered an expansion state, a state must have expanded eligibility for Medicaid through the Affordable Care Act by mid-2015.

Similar estimates were obtained after we adjusted for state trends and observed covariates. In Medicaid expansion states, the poor gained 15.0 percentage points in insurance coverage and 7.7 percentage points in having a primary care provider. The percentage of poor respondents avoiding care due to cost fell by 7.5 percentage points ( Exhibit 3 ). Households with annual incomes above $75,000 experienced much smaller changes: a 1.9-percentage-point increase in insurance coverage, a 1.9-percentage-point increase in having a primary care provider, and no change in avoiding care due to cost. Gains in access were substantially larger among people who were not college graduates, compared to those who were; renters, compared to homeowners; and the unemployed, compared to the employed.

In general, residents of nonexpansion states also had increased access. However, compared to people in expansion states, residents of nonexpansion states had smaller gains, and the distribution of benefits was less concentrated in lower socioeconomic groups. (For a comparison of crude and adjusted estimates, see Appendix Exhibits S2, S4, and S6.) ²⁷

To what extent were changes in access attributable to Medicaid expansion? In adjusted difference-in-differences models, Medicaid expansion was associated with a 2.2-percentage-point increase in insurance coverage in the full sample, after adjustment for covariates ( Exhibit 3 ) (95% confidence interval: 0.8, 3.6). The benefits of expansion were particularly large among respondents in poor households (6.3 percentage points; 95% CI: 3.2, 9.4) the unemployed (11.0 percentage points; 95% CI: 5.2, 16.8), those who were not college graduates (3.2 percentage points; 95% CI: 1.4, 5.0), and renters (2.8 percentage points; 95% CI, 0.2, 5.4). By contrast, Medicaid expansion was associated with near-zero changes in coverage among nonpoor respondents, college graduates, and the employed.

Changes in access to a primary care provider and avoiding care due to cost followed patterns similar to those for insurance coverage, although the changes were smaller. Among poor Americans, Medicaid expansion reduced the percentage without a primary care provider by 3.6 percentage points (95% CI: 0.4, 6.8) and the percentage who avoided medical care due to cost by 3.5 percentage points (95% CI: 0.4, 6.6) ( Exhibit 3 ). (For a comparison of crude and adjusted difference-in-differences estimates and coefficient standard errors, see Appendix Exhibits S2, S4, and S6.) ²⁷ A sensitivity analysis excluding states that expanded after January 1, 2014, had similar results. (For results of the sensitivity analyses, see Appendix Exhibits S8 and S9.) ²⁷

What was the impact of the ACA on socioeconomic disparities in access? In expansion states, the gap in insurance coverage between residents of poor households (with incomes less than $25,000) and higher-income households (incomes more than $75,000) fell by 46 percent between 2013 and 2015, from 31 percentage points to 17 percentage points, while in nonexpansion states the coverage gap fell by 23 percent, from 36 percentage points to 28 percentage points ( Exhibit 4 ). Income-related gaps in access to a primary care provider and avoiding care due to cost also declined more in expansion states than in nonexpansion states ( Exhibit 4 ). There were also greater reductions in health care access disparities based on education level and employment status in expansion versus nonexpansion states. (For data on both relative and absolute changes in access gaps based on different socioeconomic characteristics, as well as confidence intervals, see Appendix Exhibits S10 and S11.) ²⁷

SOURCE Authors’ analysis of data for 2011–15 from the Behavioral Risk Factor Surveillance System. NOTES The exhibit displays percent changes from 2013 to 2015 in health care access gaps between low- and high-income noninstitutionalized US adults ages 18–64. The data are stratified by whether the state expanded Medicaid. Percent changes were calculated as the access gap in 2015 divided by the access gap in 2013, minus one. All changes in access gaps were statistically significant ( $p<0.05$ ). Low income means household income of less than $25,000. High income means household income of more than $75,000. As explained in more detail in the text, to be considered an expansion state, a state must have expanded eligibility for Medicaid through the Affordable Care Act by mid-2015. Changes in access gaps by educational attainment, employment status, and homeownership are shown in online Appendix Exhibit S8 (see Note  27 in text).

Not only did access disparities fall in greater absolute terms in expansion states compared to nonexpansion states, but disparities were also smaller in expansion states in the first place (see Appendix Exhibits S2 and S11). States’ opting out of the ACA Medicaid expansion thus compounded preexisting access barriers for their poorer residents, leading to a geographic divergence in access for poor Americans. In 2013, poor residents of nonexpansion states were 22 percent (8.2 percentage points) more likely to be uninsured than poor residents of expansion states. After the ACA’s passage, this geographic disparity increased: By 2015, poor Americans were 66 percent (14.0 percentage points) more likely to be uninsured if they lived in a nonexpansion state than if they lived in an expansion state (see Appendix Exhibit S2). ²⁷

Discussion

We examined the extent to which the ACA reduced disparities in health care access across socioeconomic groups and assessed the contribution of Medicaid expansion to these trends. Americans in groups with lower socioeconomic status made substantial gains in access during the first two years of full implementation of the ACA. Medicaid expansion was responsible for about half of these gains, with the rest likely attributable to other aspects of the ACA implemented in all states in 2014, such as the insurance exchanges, federal subsidies for the purchase of insurance for people with incomes of 100–400 percent of poverty, and the individual mandate. Disparities in access narrowed significantly under the ACA, with the gap in coverage between higher- and lower-income households falling by 46 percent in Medicaid expansion states and 23 percent in nonexpansion states.

In spite of the substantial reduction in access gaps under the ACA, many Americans with household incomes under $25,000 were still without coverage in 2015: 35 percent in nonexpansion states and 21 percent in expansion states. Additionally, in 2015, many low- and middle-income Americans still reported avoiding care due to cost and said that they did not have a primary care provider. Incomplete insurance uptake might be due to factors such as unawareness of coverage options, ⁴² complicated enrollment processes, ⁴³ political attitudes toward the ACA, ⁴⁴ lack of Medicaid expansion, and the cost and low perceived value of existing plans. Understanding people’s reasons for not taking up insurance under the ACA will be important in designing policies to further reduce access gaps.

Health care access disparities in the United States far exceed those observed in other wealthy countries, which by and large guarantee some basic level of health coverage ⁴⁵ (for cross-national comparisons, see Appendix Exhibit S13). ²⁷ Reducing these disparities has been a subject of national debate since President Harry Truman proposed universal coverage in 1945. As we have shown in this analysis, the ACA substantially improved health insurance coverage and access to care for the poor and significantly reduced socioeconomic gaps in health care access in just two years. The ACA was a highly redistributive law, directing public resources, financed primarily through taxes on high-income people, to improve health care access among lower-income Americans. Those who benefited most from the ACA were those most likely to be excluded by an employer-based insurance system: the unemployed, those without a college degree, and those earning a low income. With significant numbers of Americans out of work and an increasing share of jobs not offering employer-based health coverage, there is a growing need for a robust, publicly funded insurance safety net.

President Trump and Republicans in Congress have promised to repeal and replace the ACA, with plans to reduce federal subsidies for Medicaid expansion and for low-income (but not Medicaid-eligible) insurance plans purchased on the exchanges. Such an approach is likely to widen gaps in health care access between lower-income and better-off Americans, reversing gains observed under the ACA.

Conclusion

In its first two years of full implementation, the ACA improved health care access for Americans in low-income households, people who were not college graduates, and the unemployed. The law’s Medicaid expansion was responsible for about half of these gains. The ACA was associated with a substantial (but incomplete) narrowing of socioeconomic disparities in access, particularly in states that expanded Medicaid. More research is needed to determine whether existing access gains will translate into improved health outcomes and reductions in health disparities more broadly, and to monitor future trends in access disparities in a changing policy environment.

ACKNOWLEDGMENTS

A previous version of this article was presented at the 22nd Annual National Research Service Award Research Trainees Conference, hosted by the Agency for Healthcare Research and Quality (AHRQ), Boston, Massachusetts, June 25, 2016, and at the International Health Economics Association World Congress, Boston, Massachusetts, July 7–11, 2017. Kevin Griffith received financial support from the AHRQ National Research Service Award Institutional Health Services Research Training Program (Grant No. 5 T32 HS 22242-4). Jacob Bor received financial support from the Peter T. Paul Career Development Professorship. The authors acknowledge the thoughtful feedback of James F. Burgess, Eva Dugoff, and David H. Bor on an earlier draft of this article.

NOTES

• 1 Henry J. Kaiser Family Foundation . Status of state action on the Medicaid expansion decision [Internet]. Menlo Park (CA) : KFF ; 2017 [cited 2017 Jun 19 ]. Available for download from: http://www.kff.org/health-reform/state-indicator/state-activity-around-expanding-medicaid-under-the-affordable-care-act/ Google Scholar
• 2 Mensah GA , Mokdad AH , Ford ES , Greenlund KJ , Croft JB . State of disparities in cardiovascular health in the United States . Circulation . 2005 ; 111 ( 10 ): 1233 – 41 . Crossref, Medline, Google Scholar
• 3 Pappas G , Queen S , Hadden W , Fisher G . The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and 1986 . N Engl J Med . 1993 ; 329 ( 2 ): 103 – 9 . Crossref, Medline, Google Scholar
• 4 Ward E , Jemal A , Cokkinides V , Singh GK , Cardinez C , Ghafoor A , et al. Cancer disparities by race/ethnicity and socioeconomic status . CA Cancer J Clin . 2004 ; 54 ( 2 ): 78 – 93 . Crossref, Medline, Google Scholar
• 5 Baicker K , Taubman SL , Allen HL , Bernstein M , Gruber JH , Newhouse JP , et al. The Oregon experiment—effects of Medicaid on clinical outcomes . N Engl J Med . 2013 ; 368 ( 18 ): 1713 – 22 . Crossref, Medline, Google Scholar
• 6 Institute of Medicine . America’s uninsured crisis: consequences for health and health care . Washington (DC) : National Academies Press ; 2009 . p. 214 . Google Scholar
• 7 Smith JC , Medalia C . Health insurance coverage in the United States: 2013 [Internet]. Washington (DC) : Census Bureau ; 2014 Sep [cited 2017 Jun 19 ]. (Current Population Report No. P60-250). Available from: https://www.census.gov/content/dam/Census/library/publications/2014/demo/p60-250.pdf Google Scholar
• 8 Wisk LE , Witt WP . Predictors of delayed or forgone needed health care for families with children . Pediatrics . 2012 ; 130 ( 6 ): 1027 – 37 . Crossref, Medline, Google Scholar
• 9 Hafner-Eaton C . Physician utilization disparities between the uninsured and insured. Comparisons of the chronically ill, acutely ill, and well nonelderly populations . JAMA . 1993 ; 269 ( 6 ): 787 – 92 . Crossref, Medline, Google Scholar
• 10 Swan J , Breen N , Coates RJ , Rimer BK , Lee NC . Progress in cancer screening practices in the United States: results from the 2000 National Health Interview Survey . Cancer . 2003 ; 97 ( 6 ): 1528 – 40 . Crossref, Medline, Google Scholar
• 11 Dickman SL , Woolhandler S , Bor J , McCormick D , Bor DH , Himmelstein DU . Health spending for low-, middle-, and high-income Americans, 1963–2012 . Health Aff (Millwood) . 2016 ; 35 ( 7 ): 1189 – 96 . Go to the article, Google Scholar
• 12 Lillie-Blanton M , Hoffman C . The role of health insurance coverage in reducing racial/ethnic disparities in health care . Health Aff (Millwood) . 2005 ; 24 ( 2 ): 398 – 408 . Go to the article, Google Scholar
• 13 Chetty R , Stepner M , Abraham S , Lin S , Scuderi B , Turner N , et al. The association between income and life expectancy in the United States, 2001–2014 . JAMA . 2016 ; 315 ( 16 ): 1750 – 66 . Crossref, Medline, Google Scholar
• 14 Bor J , Cohen GH , Galea S . Population health in an era of rising income inequality: USA, 1980–2015 . Lancet . 2017 ; 389 ( 10077 ): 1475 – 90 . Crossref, Medline, Google Scholar
• 15 Smith JC , Medalia C . Health insurance coverage in the United States: 2014 [Internet]. Washington (DC) : Census Bureau ; 2015 Sep [cited 2017 Jun 19 ]. (Current Population Report No. P60-253). Available from: https://www.census.gov/content/dam/Census/library/publications/2015/demo/p60-253.pdf Google Scholar
• 16 Chen J , Vargas-Bustamante A , Mortensen K , Ortega AN . Racial and ethnic disparities in health care access and utilization under the Affordable Care Act . Med Care . 2016 ; 54 ( 2 ): 140 – 6 . Crossref, Medline, Google Scholar
• 17 Torres H , Poorman E , Tadepalli U , Schoettler C , Fung CH , Mushero N , et al. Coverage and access for Americans with chronic disease under the Affordable Care Act: a quasi-experimental study . Ann Intern Med . 2017 ; 166 ( 7 ): 472 – 9 . Crossref, Medline, Google Scholar
• 18 Sommers BD , Gunja MZ , Finegold K , Musco T . Changes in self-reported insurance coverage, access to care, and health under the Affordable Care Act . JAMA . 2015 ; 314 ( 4 ): 366 – 74 . Crossref, Medline, Google Scholar
• 19 Sommers BD , Musco T , Finegold K , Gunja MZ , Burke A , McDowell AM . Health reform and changes in health insurance coverage in 2014 . N Engl J Med . 2014 ; 371 ( 9 ): 867 – 74 . Crossref, Medline, Google Scholar
• 20 Benitez JA , Creel L , Jennings J . Kentucky’s Medicaid expansion showing early promise on coverage and access to care . Health Aff (Millwood) . 2016 ; 35 ( 3 ): 528 – 34 . Go to the article, Google Scholar
• 21 Sommers BD , Blendon RJ , Orav EJ , Epstein AM . Changes in utilization and health among low-income adults after Medicaid expansion or expanded private insurance . JAMA Intern Med . 2016 ; 176 ( 10 ): 1501 – 9 . Crossref, Medline, Google Scholar
• 22 Gaffney A , McCormick D . The Affordable Care Act: implications for health-care equity . Lancet . 2017 ; 389 ( 10077 ): 1442 – 52 . Crossref, Medline, Google Scholar
• 23 McMorrow S , Kenney GM , Long SK , Anderson N . Uninsurance among young adults continues to decline, particularly in Medicaid expansion states . Health Aff (Millwood) . 2015 ; 34 ( 4 ): 616 – 20 . Go to the article, Google Scholar
• 24 Shartzer A , Long SK , Anderson N . Access to care and affordability have improved following Affordable Care Act implementation; problems remain . Health Aff (Millwood) . 2016 ; 35 ( 1 ): 161 – 8 . Go to the article, Google Scholar
• 25 Wherry LR , Miller S . Early coverage, access, utilization, and health effects associated with the Affordable Care Act Medicaid expansions: a quasi-experimental study . Ann Intern Med . 2016 ; 164 ( 12 ): 795 – 803 . Crossref, Medline, Google Scholar
• 26 Miller S , Wherry LR . Health and access to care during the first 2 years of the ACA Medicaid expansions . N Engl J Med . 2017 ; 376 ( 10 ): 947 – 56 . Crossref, Medline, Google Scholar
• 27 To access the Appendix, click on the Appendix link in the box to the right of the article online.
• 28 Henry J. Kaiser Family Foundation . Medicaid income eligibility limits for other non-disabled adults, 2011–2017 [Internet]. Menlo Park (CA) : KFF ; 2017 [cited 2017 Jun 19 ]. Available from: http://kff.org/medicaid/state-indicator/medicaid-income-eligibility-limits-for-other-non-disabled-adults/ Google Scholar
• 29 Pulos V . MassHealth advocacy guide—2012 [Internet]. Boston (MA) : MassLegal Services ; 2012 Apr 4 . Part 10, Eligibility criteria for uninsured adults in Commonwealth Care; [cited 2017 Jun 19 ]. Available for download from: https://www.masslegalservices.org/content/masshealth-advocacy-guide-2012 Google Scholar
• 30 Tucker S . Medicaid/PAC update [Internet]. Baltimore (MD) : Maryland Department of Health and Mental Hygiene, Office of Health Services ; 2010 May 14 [cited 2017 Jul 5 ]. Available from: http://madc.homestead.com/Susan_Tucker_Presentation.pdf Google Scholar
• 31 Walker AK . Understanding Medicaid expansion under health reform. Baltimore Sun [serial on the Internet]. 2013 Sep 6 [cited 2017 Jun 19 ]. Available from: http://www.baltimoresun.com/health/blog/bs-hs-reform-question-medicaid-20130906-story.html Google Scholar
• 32 Pierannunzi C , Hu SS , Balluz L . A systematic review of publications assessing reliability and validity of the Behavioral Risk Factor Surveillance System (BRFSS), 2004–2011 . BMC Med Res Methodol . 2013 ; 13 ( 1 ): 49 . Crossref, Medline, Google Scholar
• 33 Centers for Disease Control and Prevention . Behavioral Risk Factor Surveillance System: weighting BRFSS data: BRFSS 2013 [Internet]. Atlanta (GA) : CDC ; 2013 [cited 2017 Jun 19 ]. Available from: https://www.cdc.gov/brfss/annual_data/2013/pdf/weighting_data.pdf Google Scholar
• 34 Bertrand M , Duflo E , Mullainathan S . How much should we trust differences-in-differences estimates? Q J Econ . 2004 ; 119 ( 1 ): 249 – 75 . Crossref, Google Scholar
• 35 Hill I , Wilkinson M , Courtot B . The launch of the Affordable Care Act in selected states: outreach, education, and enrollment assistance [Internet]. Washington (DC) : Urban Institute ; 2014 Mar [cited 2017 Jun 19 ]. Available from: http://www.urban.org/sites/default/files/publication/22341/413039-The-Launch-of-the-Affordable-Care-Act-in-Eight-States-Outreach-Education-and-Enrollment-Assistance.pdf Google Scholar
• 36 Centers for Disease Control and Prevention . Behavioral Risk Factor Surveillance System: 2014 summary data quality report [Internet]. Atlanta (GA) : CDC ; 2015 Jul 29 [cited 2017 Jun 19 ]. Available from: https://www.cdc.gov/brfss/annual_data/2014/pdf/2014_dqr.pdf Google Scholar
• 37 Executive Office of the President of the United States . The economic record of the Obama Administration: reforming the health care system [Internet]. Washington (DC) : The Office ; 2016 Dec [cited 2017 Jun 19 ]. Available from: https://obamawhitehouse.archives.gov/sites/default/files/page/files/20161213_cea_record_healh_care_reform.pdf Google Scholar
• 38 Obama B . United States health care reform: progress to date and next steps . JAMA . 2016 ; 316 ( 5 ): 525 – 32 . Crossref, Medline, Google Scholar
• 39 Bauchner H . The Affordable Care Act and the future of US health care . JAMA . 2016 ; 316 ( 5 ): 492 – 3 . Crossref, Medline, Google Scholar
• 40 Lipton BJ , Wherry LR , Miller S , Kenney GM , Decker S . Previous Medicaid expansion may have had lasting positive effects on oral health of non-Hispanic black children . Health Aff (Millwood) . 2016 ; 35 ( 12 ): 2249 – 58 . Go to the article, Google Scholar
• 41 Sommers BD , Maylone B , Blendon RJ , Orav EJ , Epstein AM . Three-year impacts of the Affordable Care Act: improved medical care and health among low-income adults . Health Aff (Millwood) . 2017 ; 36 ( 6 ): 1119 – 28 . Go to the article, Google Scholar
• 42 Sommers BD , Maylone B , Nguyen KH , Blendon RJ , Epstein AM . The impact of state policies on ACA applications and enrollment among low-income adults in Arkansas, Kentucky, and Texas . Health Aff (Millwood) . 2015 ; 34 ( 6 ): 1010 – 8 . Go to the article, Google Scholar
• 43 Baicker K , Congdon WJ , Mullainathan S . Health insurance coverage and take-up: lessons from behavioral economics . Milbank Q . 2012 ; 90 ( 1 ): 107 – 34 . Crossref, Medline, Google Scholar
• 44 Henry J. Kaiser Family Foundation . Kaiser health tracking poll: the public’s views on the ACA [Internet]. Menlo Park (CA) : KFF ; 2017 May 31 [cited 2017 Jun 20 ]. Available from: http://www.kff.org/interactive/kaiser-health-tracking-poll-the-publics-views-on-the-aca/#?response=Favorable--Unfavorable&aRange=twoYear Google Scholar
• 45 Davis K , Stremikis K , Squires D , Schoen C . Mirror, mirror on the wall. How the performance of the US health care system compares internationally . New York (NY) : Commonwealth Fund ; 2014 . Google Scholar