Affordable Care Act Medicaid Expansion Reduced Uninsured Hospital Stays In 2014

In their first year, the main coverage provisions of the Affordable Care Act (ACA)—related to the health insurance Marketplaces and expansions of eligibility for Medicaid—resulted in up to seventeen million Americans’ gaining coverage. ^1–4 Early analyses suggest that coverage expansion has increased access to physicians and reduced cost-related barriers to care, with the largest effects occurring in the twenty-nine states (including the District of Columbia) that had expanded Medicaid by the first quarter of 2015. ^4,5

In addition to the benefits of coverage for the newly insured, the prospect that reducing the number of uninsured patients would also reduce hospitals’ burden of uncompensated care figured importantly in the debate over the ACA and persists in state-level discussions concerning Medicaid expansion. ^6,7 To shed light on whether coverage expansions are delivering on this promise, we present newly available data on trends in payer mix for non-Medicare adult inpatient hospital stays from HCUP Fast Stats, a new online database query tool from the Agency for Healthcare Research and Quality (AHRQ). ⁸

States that expanded eligibility for Medicaid in 2014 (Arizona, California, Colorado, Hawaii, Iowa, Kentucky, Minnesota, New Jersey, and New York) saw dramatic decreases in uninsured hospital stays and increases in Medicaid-covered stays, while those that did not (Florida, Georgia, Indiana, Missouri, Virginia, and Wisconsin) experienced very little change in the mix of payers for inpatient care ( Exhibit 1 ). Indiana expanded its Medicaid program in 2015. Results are based on fifteen states for which data were available through at least the second quarter of 2014.

SOURCE Authors’ analysis of discharge data from the Agency for Healthcare Research and Quality’s HCUP Fast Stats (see Note   8 in text). NOTES The data come from nine states that expanded eligibility for Medicaid (Arizona, California, Colorado, Hawaii, Iowa, Kentucky, Minnesota, New Jersey, and New York) and six states that did not (Florida, Georgia, Indiana, Missouri, Virginia, and Wisconsin). Data are weighted by total non-Medicare hospital discharges. The orange line indicates the beginning of the implementation of the Affordable Care Act’s major coverage provisions, including Medicaid expansion.

These results build on previous analyses of the early experiences of several states and of five large for-profit hospital chains, which suggested that coverage expansions shifted the mix of payers for inpatient care toward Medicaid and away from self-pay, reducing uncompensated care. ^9–11 Our analysis also makes it possible to disentangle the relative importance of Medicaid expansion versus the private Marketplace expansion in driving changes in payer mix. This analysis provides the best evidence to date on how ACA coverage expansions affected the mix of payers for hospitalized patients, and what states that so far have chosen not to expand Medicaid might expect to experience should they decide to embrace expansion in the future.

Study Data And Methods

HCUP Fast Stats is a new online database query tool maintained by AHRQ that reports adult hospital discharges by calendar quarter at the state level. ⁸ The underlying data are drawn from state hospital discharge databases participating in the Healthcare Cost and Utilization Project (HCUP) that contain information on all discharges in the state and cover more than 95 percent of inpatient hospitalizations in each state. HCUP Fast Stats provides state-level data aggregated by patient age and primary expected source of payment (Medicaid, private insurance, Medicare, and no insurance).

Forty-one states participate in HCUP Fast Stats, and data through at least the second quarter of 2014 are available for sixteen of these states. We excluded one of those states—Michigan—from our analysis because its Medicaid expansion did not take effect until the second quarter of 2014. Collectively, the remaining fifteen states represent 54 percent of the US population. Our analytic sample consisted of quarterly observations from these states, from the first quarter of 2009 through the second quarter of 2014.

To assess the impact of the ACA’s coverage provisions on hospital payer mix, we divided states into those that did and those that did not expand Medicaid, and we compared changes before and after coverage expansions. Underlying payer-mix data are available in online Appendix Exhibit 1. ¹² We used multivariate linear regression to compare changes in payer mix over time between states that did and those that did not expand Medicaid in a difference-in-differences analysis, controlling for state-level demographic and economic characteristics.

We also present state-specific changes in payer mix between the third quarter of 2013 and the second quarter of 2014. The state-specific results show that the aggregate differences between expansion and nonexpansion states evident in Exhibit 1 were not driven by the experiences of one or two large states in each group; instead, they reflect trends that were evident across most states within each group.

Study Results

Several important facts are evident from Exhibit 1 . First, the proportions of Medicaid and uninsured inpatients were gradually increasing during the years leading up to 2014, while the proportion of inpatients with private coverage was gradually decreasing. This mirrors well-documented trends in the insurance coverage of the general population. ¹³

Second, before the ACA coverage expansions, states that later expanded Medicaid had higher Medicaid shares and lower uninsured shares, compared to nonexpansion states. However, trends in the inpatient payer mix were quite similar in expansion and nonexpansion states.

Third, and most striking, the data show a sharp break in the trends in Medicaid and uninsured shares in the first two quarters of 2014 for expansion states only. Between the third quarter of 2013 and the second quarter of 2014, expansion states experienced a 7-percentage-point jump in the Medicaid share and a 6-percentage-point drop in the uninsured share. These differences represent a 20 percent increase and a 50 percent decrease in Medicaid and uninsured discharges, respectively. Meanwhile, in nonexpansion states, changes in Medicaid and uninsured discharges were small—less than 1 percentage point—and not significant.

Exhibit 2 summarizes the levels of and changes in payer mix between the period before expansion, defined as the first quarter of 2009 through the third quarter of 2013, and the latest period after expansion for which data were available, defined as the second quarter of 2014, for the two groups of states. We chose to include the fourth quarter of 2013 in the period after expansion because previous work has found that Medicaid coverage began to increase during the open enrollment period. ^2,4

Exhibit 2 also presents simple difference-in-differences estimates of the effect of Medicaid expansion, including adjusted differences-in-differences based on multivariate regression analyses that controlled for economic and demographic differences across individual states (full results are available in Appendix Exhibit 2). ¹² The results from this more rigorous statistical analysis confirm the significant drop in the uninsured and the significant increase in the Medicaid discharges in expansion—but not in nonexpansion—states following the coverage expansions in 2014 ( Exhibit 1 ).

Finally, we saw no significant change in the share of inpatients with private insurance in expansion states relative to nonexpansion states ( Exhibit 2 ). This should not be taken to mean that the ACA did not expand private coverage. On the contrary, previous work suggests that it has. ^2–4 A more likely explanation is that the majority of uninsured individuals who were sick enough to need hospital care in 2013 gained Medicaid, instead of private coverage, in 2014 because they had very little income. In addition, because of the initial problems with the Marketplace enrollment websites, much of the enrollment in private health plans occurred at the end of the first quarter of 2014.

For a closer look at what happened in each of the study states, we also present unadjusted state-specific changes between the third quarter of 2013 and the second quarter of 2014 in the fraction of hospital stays that were uninsured ( Exhibit 3 ), covered by Medicaid ( Exhibit 4 ), and covered by private insurance ( Exhibit 5 ). A formal statistical analysis of how trends in payer mix in each state changed in the period after expansion is presented in Appendix Exhibit 3. ¹² Because that analysis largely confirms the observations seen in the exhibits, we focus here on the exhibits.

SOURCE Authors’ analysis of discharge data from the Agency for Healthcare Research and Quality’s HCUP Fast Stats (see Note  8 in text).

SOURCE Authors’ analysis of discharge data from the Agency for Healthcare Research and Quality’s HCUP Fast Stats (see Note  8 in text).

SOURCE Authors’ analysis of discharge data from the Agency for Healthcare Research and Quality’s HCUP Fast Stats (see Note  8 in text).

Uninsured hospitalizations declined in the majority of states, but the declines were much larger in expansion than in nonexpansion states ( Exhibit 3 ). This result would be expected given the trends shown in Exhibit 1 . The decrease in uninsured discharges was especially pronounced in Kentucky, where they fell by 13.5 percentage points.

Medicaid stays increased sharply in all nine expansion states except Minnesota, which had already expanded Medicaid to low-income adults in 2011 ( Exhibit 4 ). At that time the Medicare share of discharges increased sharply in Minnesota (data not shown). In contrast, among nonexpansion states, only Wisconsin experienced a meaningful increase in Medicaid discharges. This was likely because all adults in Wisconsin with incomes of up to 100 percent of the federal poverty level were already eligible for Medicaid. We speculate that this reflects a “welcome mat” effect, in which previously eligible but unenrolled individuals signed up for coverage in 2014.

Although the aggregate trend in the share of hospital stays covered by private insurance showed little change in 2014 ( Exhibit 1 ), some states did experience increases in private coverage for hospital stays ( Exhibit 5 ). For example, Virginia and Florida—both nonexpansion states—experienced an uptick in private-payer share in 2014 (for a table of regression-adjusted state-specific changes, see the Appendix). ¹² In contrast, Georgia, another nonexpansion state, saw its share of private discharges fall by a little more than 3 percentage points. This heterogeneity illustrates the importance of considering the unique circumstances of each state, a consideration with which state policy makers are already likely well acquainted.

Discussion

Our analysis used discharge data through the second quarter of 2014. Our findings suggest that the ACA resulted in immediate changes in payer mix for hospitals in states that expanded Medicaid eligibility, with reductions in patients having no expected source of payment and increases in patients covered by Medicaid. These changes should reduce hospitals’ burden of uncompensated care.

However, with only half a year of post-ACA data available, our results were limited, in that they may not reflect the experience of hospitals in all of 2014 or in 2015. A definitive analysis of this issue awaits population-level data on both insurance coverage and health care utilization. In addition, a complete analysis of how changes in payer mix affect the amount of uncompensated care provided by hospitals will require Medicare cost report data for fiscal years 2014 and 2015, which are not yet available.

Conclusion

Our findings underscore the significant benefits of Medicaid expansion not only for low-income adults, but also for the hospitals that serve this population. Understanding the impact of Medicaid expansion on hospitals will become even more important as we approach 2017, when hospitals in all states will begin facing increasingly large annual cuts in disproportionate-share hospital payments that subsidize the cost of uncompensated care. ¹⁴

ACKNOWLEDGMENTS

The authors thank Anne Elixhauser and Claudia Steiner for helpful comments. Helen Levy acknowledges financial support from the National Institute on Aging (Grant No. NIA K01AG034232).

NOTES

• 1 Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation . Health insurance coverage and the Affordable Care Act [Internet]. Washington (DC) : HHS ; 2015 May 5 [cited 2015 Nov 30 ]. Available from: http://aspe.hhs.gov/sites/default/files/pdf/139211/ib_uninsured_change.pdf Google Scholar
• 2 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
• 3 Carman KG , Eibner C , Paddock SM . Trends in health insurance enrollment, 2013–15 . Health Aff (Millwood) . 2015 ; 34 ( 6 ): 1044 – 8 . Go to the article, Google Scholar
• 4 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
• 5 As of March 2015 twenty-nine states (including the District of Columbia) had expanded Medicaid eligibility under the ACA. As of December 2015 thirty-one states (including the District of Columbia) had decided to expand Medicaid. Henry J. Kaiser Family Foundation . Status of state action on the Medicaid expansion decision [Internet]. Menlo Park (CA) : KFF ; [cited 2015 Dec 7 ]. Available from: http://kff.org/health-reform/state-indicator/state-activity-around-expanding-medicaid-under-the-affordable-care-act/ Google Scholar
• 6 Ollove M . Hospitals lobby hard for Medicaid expansion [Internet]. Philadelphia (PA) : Pew Charitable Trusts ; 2013 Apr 17 [cited 2015 Nov 30 ]. Available from: http://www.pewtrusts.org/en/research-and-analysis/blogs/stateline/2013/04/17/hospitals-lobby-hard-for-medicaid-expansion Google Scholar
• 7 Barnes J . 2014 . Virginia hospitals are lobbying for an expansion of Medicaid . Washington Post . 2014 Feb 5 . Google Scholar
• 8 Agency for Healthcare Research and Quality . HCUP Fast Stats—effect of Medicaid expansion on hospital use [Internet]. Rockville (MD) : AHRQ ; [last modified 2015 Jun 5; cited 2015 Nov 30 ]. Available from: https://www.hcup-us.ahrq.gov/faststats/statepayer/states.jsp Google Scholar
• 9 Nikpay S , Buchmueller T , Levy H . Early Medicaid expansion in Connecticut stemmed the growth in hospital uncompensated care . Health Aff (Millwood) . 2015 ; 34 ( 7 ): 1170 – 9 . Go to the article, Google Scholar
• 10 DeLeire T , Joynt K , McDonald R . Impact of insurance coverage and hospital uncompensated care costs in 2014 [Internet]. Washington (DC) : Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation ; 2014 Sep 24 [cited 2015 Nov 30 ]. (ASPE Issue Brief). Available from: http://aspe.hhs.gov/sites/default/files/pdf/77061/ib_UncompensatedCare.pdf Google Scholar
• 11 Garthwaite C , Gross T , Notowidigdo MJ . Hospitals as insurers of last resort [Internet]. Unpublished paper. 2015 Jan [cited 2015 Nov 30 ]. Available from: http://faculty.wcas.northwestern.edu/noto/research/GGN_Hospitals%20as%20Insurers%20of%20Last%20Resort_jan2015.pdf Google Scholar
• 12 To access the Appendix, click on the Appendix link in the box to the right of the article online.
• 13 Denavas-Walt C , Proctor BD . Income and poverty in the United States: 2014 [Internet]. Washington (DC) : Census Bureau ; 2015 Sep [cited 2015 Dec 7 ]. Available from: https://www.census.gov/content/dam/Census/library/publications/2015/demo/p60-252.pdf Google Scholar
• 14 Millman J . 2014. Health care’s $85 billion challenge—uncompensated care in the Obamacare age . Washington Post Wonkblog [blog on the Internet]. 2014 May 5 [cited 2015 Nov 30 ]. Available from: https://www.washingtonpost.com/news/wonk/wp/2014/05/05/health-cares-85-billion-challenge-uncompensated-care-in-the-obamacare-age/ Google Scholar