The Effects of Medicaid Expansion under the ACA: Findings from a Literature Review
Note: This is not the most recent version of this paper. An updated version is available here.
Research on the effects of Medicaid expansions under the Affordable Care Act (ACA) can help increase understanding of how the ACA has impacted coverage; access to care, utilization, and health outcomes; and various economic outcomes, including state budgets, the payer mix for hospitals and clinics, and the employment and labor market. These findings also may inform ongoing debates surrounding the Medicaid expansion. This brief reviews and summarizes findings from a total of 61 studies of the impact of state Medicaid expansions under the ACA.
Prior to implementation of state Medicaid expansions in 2014, many analyses studied the potential effects of expansion, often using data from states that implemented early (pre-2014) expansions of Medicaid eligibility. A previous Kaiser Commission on Medicaid and the Uninsured brief titled, “The Role of Medicaid in State Economies and the ACA” summarized results from many of those early studies. This brief focuses solely on the effects of Medicaid expansion that states have experienced following ACA expansion implementation in 2014. The specific criteria used for the selection of studies included in this literature review are described in the methods section below. The brief provides an overview of the key themes and findings across the large body of literature on Medicaid expansion’s effects, and more details on the major findings from each individual study are available in the appendix tables.
Studies were selected for this literature review based on several specific criteria, all intended to preserve this project’s focus on direct effects of Medicaid expansion under the ACA. This review includes only studies conducted between January 2014 and May 2016 using data from 2014 or later, meaning that studies on the effects of early expansions prior to implementation of the major ACA coverage provisions in 2014 are excluded. The review also includes only studies that report specific impacts of the ACA Medicaid expansion—findings on impacts of ACA coverage expansions generally (not specific to Medicaid expansion alone) were not included. Other types of studies or findings that were excluded from this analysis are those published by advocacy organizations and those investigating potential effects of expansion in states that have not (or had not, at the time of the study) expanded Medicaid.
This review draws on a range of types of publications, including peer-reviewed journals, free-standing reports, government reports, and white papers. Studies were identified primarily through ProQuest, PubMed, general web searches, and searches through the websites of specific journals or organizations. While a range of search methods were used in an attempt to be as exhaustive as possible, there is still a chance that some studies that meet the inclusion criteria were missed.
In both the brief and the appendix tables, study findings are separated into three broad categories: (1) expansion’s impact on coverage, including changes in Medicaid enrollment and the size of the uninsured population, (2) expansion’s impact on access to care, utilization of services, and health outcomes, and (3) expansion’s impact on economic outcomes, including state budgets and economies, the payer mix for hospitals and clinics, and measures related to employment and the labor market.
Impacts on Coverage (Appendix Table A)
States expanding their Medicaid programs under the ACA have seen large increases in Medicaid enrollment.1,2,3,4,5 Medicaid enrollment grew substantially in expansion states following the first open enrollment period (which began in October 2013), with enrollment exceeding expectations in many states.6,7,8 The Medicaid enrollment growth in expansion states continued through the second open enrollment period (which ended in early 2015)9 and into early 2016,10 although the pace of growth slowed in some states.11 In comparison, non-expansion states experienced slower enrollment growth during the post-January 2014 period.12,13,14,15,16
Medicaid coverage gains have been driven by enrollment of adults made newly eligible for Medicaid in expansion states as well as enrollment growth among individuals who were previously eligible for but not enrolled in Medicaid17,18 One study found that the enrollment growth among those previously eligible was primarily driven by ACA changes, such as the streamlining and simplifying of Medicaid enrollment processes that occurred in all states regardless of expansion decisions as well as broader outreach efforts.19 While coverage gains have been substantial in expansion states, the share of the remaining uninsured population that is eligible for Medicaid coverage is larger in expansion states than in non-expansion states, reflecting expansion states’ broader eligibility levels.20
Many studies show that Medicaid expansion states experienced large reductions in uninsured rates that significantly exceed those in non-expansion states.21,22,23,24,25,26,27,28,29,30,31 Beginning in 2014, numerous nationwide and state-specific studies show sharp declines in uninsured rates. These analyses reveal that uninsured rates (overall and among low-income populations specifically) dropped nearly immediately in expansion states following implementation of the ACA’s coverage provisions32,33,34,35 and continued to decline through the end of 201436,37,38 and in 2015.39 The sharp declines in uninsured rates among the low-income population in expansion states are widely attributed to gains in Medicaid coverage.40,41,42,43,44 Uninsured rates among the low-income population in non-expansion states declined somewhat as well, in part as a result of the availability of ACA subsidies for private insurance to those with incomes above poverty and increased outreach and enrollment efforts surrounding the ACA in all states. Yet these reductions in non-expansion states were far more limited than the substantial declines observed in expansion states.45,46,47,48,49 Studies find similar coverage patterns—with larger coverage gains in expansion versus non-expansion states—among young adults,50 mothers,51 and childless adults with incomes under 100% of the Federal Poverty Level (FPL).52
A recent study found that, on average, implementation of the ACA with the Medicaid expansion increased coverage by 5.9 percentage points in 2014, while ACA implementation without the Medicaid expansion increased coverage by 3.0 percentage points. As such, the authors conclude that Medicaid expansion contributed to a 2.9 percentage point increase in coverage.53
Arkansas implemented an alternative Medicaid expansion, called the Private Option, under which the state uses Medicaid funds to purchase private insurance for newly eligible adults in the Marketplace. Studies of Arkansas find that the private option led to coverage gains that are consistent with the experience of states that implemented a traditional expansion.54,55,56
Community health centers in Medicaid expansion states have seen gains in coverage among their patients.57,58,59,60,61 Health centers that have historically served uninsured populations in expansion states reported increases in their share of patients with coverage, particularly Medicaid coverage. Clinics in non-expansion states, however, report much smaller changes in the share of patients they serve with coverage, since many of their patients fall into the coverage gap.62,63 Similarly, clinics serving individuals who are homeless as well as migrant health centers in expansion states experienced larger coverage gains compared to those in non-expansion states.64,65
Research suggests that the Medicaid expansion is associated with reductions in coverage disparities.66,67,68 One study found that the ACA with Medicaid expansion reduced the difference in uninsured rates between the lowest income (under 138% FPL) and highest income (over 400% FPL) groups by 8.7 percentage points, or 27%, in 2014. The ACA without the Medicaid expansion only lowered this gap by 11%.69 The same study also estimated that the ACA with the Medicaid expansion lowered the disparity in uninsured rates between Whites and Non-Whites by 2.2 percentage points (or 14%) in 2014, whereas the ACA without the Medicaid expansion actually increased this disparity.70 A separate study in Oregon found that among low-income patients who obtained community health center visits in both the pre- and post-ACA expansion periods, disparities in public insurance coverage in the pre-ACA period between Spanish-preferring Latinos, English-preferring Latinos, and non-Hispanic Whites were eliminated in the post-ACA period.71
Impacts on Access to Care, Utilization, And Health Outcomes
(Appendix Table B)
Several studies find larger improvements in measures of access to care in expansion states compared to non-expansion states. These improvements include increases in the share of individuals that have a usual source of care72,73,74 and are able to easily access medications75,76 as well as reductions in the share who lack a personal physician77 and who report problems accessing care.78 However, some studies found no statistically significant changes due to expansion in one or more access measures.79,80,81
Research suggests that Medicaid expansion improves the affordability of care and financial security among the low-income population. Several studies show that expansion states have experienced greater reductions in unmet medical need because of cost than non-expansion states.82,83,84,85 Although a few studies did not identify statistically significant differences in changes in unmet medical need due to cost between expansion and non-expansion states,86,87,88,89 some of these findings may have been affected by study design or data limitations. One of these analyses used a measure that asked respondents about unmet need due to cost in the previous 12 months, which was a period that included some months before the Medicaid expansion for most respondents.90 Another study compared unmet need due to cost in expansion and non-expansion states among all nonelderly adults rather than among the low-income group specifically, despite finding a large decline across all states in unmet need due to cost for the low-income population.91 Looking at another dimension of health care affordability, multiple studies found larger declines in trouble paying medical bills in expansion states relative to non-expansion states.92,93 A separate study found that, among those residing in areas with high shares of low-income, uninsured individuals, Medicaid expansion significantly reduced the number of unpaid bills and the amount of debt sent to third-party collection agencies.94
Studies also find changes in utilization following the Medicaid expansion. Studies show larger increases in utilization of some services in expansion states compared to non-expansion states, including increases in overnight hospital stays among low-income, nonelderly adults95 as well as increases in physician visits.96,97,98,99 One study found that expansion increased utilization of certain types of preventive care, including dental visits, breast exams, and mammograms, among childless adults below 100% FPL.100 State-specific research also supports the idea that expansion increases utilization of some health services—a study in Kentucky found that in 2014, expansion enrollees utilized common preventive care services (such as medication monitoring and cholesterol screening services) at higher rates than traditional Medicaid enrollees in the state.101 In addition, one study found that there was a larger increase in the share of individuals with chronic conditions who obtained regular care in the two expansion states included in the study (Kentucky and Arkansas) compared to a non-expansion state (Texas).102 Another national study found a larger increase in the number of Medicaid prescriptions filled in expansion states compared to non-expansion states.103
Two analyses did not find significant impacts of expansion on office visits,104 specialist visits,105 emergency department visits,106,107 and overnight hospitalizations.108 The authors of one of these studies suggested that their sample size may have been too small to detect utilization changes after only one year,109 and the second study’s authors similarly noted that if changes in utilization take more than one year to materialize, it may have been too soon to observe the full effect of expansion.110
Several studies have examined access and utilization among enrollees in Arkansas where the Medicaid expansion was implemented through a waiver. One study examined how several measures of access compared between Arkansas, which implemented its expansion through a “private option” waiver, and Kentucky, which implemented a traditional expansion. It did not find any significant differences in most measures, although it did show greater reduction in trouble paying medical bills in Kentucky compared to Arkansas.111 Another study of Arkansas’ Private Option Medicaid expansion found that Private Option enrollees, who are enrolled in managed care plans, have access to substantially more providers than traditional Medicaid enrollees, who are enrolled in fee-for-service arrangements, because they have access to broader networks.112 This study also found that Private Option beneficiaries utilized emergency department services at a higher rate than traditional Medicaid beneficiaries, despite being a healthier population.113 The study partially attributes higher emergency use to a lack of understanding of how to use the health care system by newly insured individuals and to a lack of incentives for using more appropriate care.114 Early reports indicated that Private Option beneficiaries were protected from upfront payments for premiums and cost-sharing that exceed Medicaid limits, while access to the two wrap-around benefits required by Medicaid but not covered in the Marketplace—non-emergency medical transportation and Early, Periodic Screening, Diagnosis, and Treatment (EPSDT) for 19 and 20-year-olds—was more mixed.115
Research shows that providers have experienced increases in Medicaid patient volume following the expansion, and results are mixed with regard to provider capacity to meet increased demands for care. Studies find that providers in states that expanded Medicaid are experiencing increases in their number and share of Medicaid patients.116,117,118 Some providers have expanded capacity to meet increased demand for care following expansion. For example, health centers in expansion states were significantly more likely than those in non-expansion states to report having expanded their capacity for dental services (37% vs. 31%) and mental health services (42% vs. 35%) since the start of 2014.119 State officials as well as providers serving the homeless population in expansion states reported that Medicaid expansion increased the availability of behavioral health treatment.120,121 In Kentucky, more than 300 new behavioral health providers enrolled in Medicaid and at least 13,000 individuals with a substance use disorder received related treatment services during the state’s first year of expansion.122 A study of the early effects of the Medicaid expansion in Michigan found that access to appointments for new patients with Medicaid improved and wait times in clinics that accepted Medicaid remained stable following expansion implementation.123
However, a few studies suggest that professional shortages may exist in some areas124 and may be especially common for behavioral health and other specialty service areas.125,126 Additionally, one nationwide study found that health centers in expansion states were more likely to report increased wait times for appointments in 2014 compared to non-expansion states.127 Given these mixed findings and the likelihood that these access challenges vary by state and region, further research and better access measures are needed to understand where these challenges occur as well as the extent of their impact on overall access to care and health outcomes for the Medicaid population.
Research suggests that improved access and utilization may be contributing to increased diagnoses of some chronic conditions. Two studies found significant increases in the rates of diagnosis of diabetes among the low-income population under state Medicaid expansions.128,129 One of these studies also found increased diagnoses of high cholesterol but no significant change in hypertension diagnoses.130
More research is needed to fully determine the impact of state Medicaid expansions on health outcomes for the low-income population. The expansion of Medicaid coverage to a larger population and the associated access and affordability benefits described above—including improved access to primary and behavioral health services as well as decreased cost barriers to receiving needed care—may contribute to improved health outcomes for the low-income population. In a study of providers serving individuals who are homeless, providers stressed that Medicaid coverage gains under expansion have allowed patients to access a broader array of services and treatment options that are improving their health outcomes—providers described some instances of individuals receiving life-saving or life-changing surgeries or treatments that they could not obtain while uninsured prior to expansion.131 Additionally, one recent study of childless, nonelderly adults with incomes below poverty found that Medicaid expansions resulted in modest improvements in self-rated health and decreases in the number of work days missed due to poor health.132
However, multiple studies on newly-insured Medicaid enrollees’ self-reported health status before and after expansion implementation have not documented significant changes in health outcomes.133,134 The authors of one study suggested that increased contact with health care professionals as a result of expansion and individuals’ subsequent improved knowledge about their health conditions may negatively affect their perceived health in the short term.135 Moreover, given that it may take time for measurable changes in health outcomes to occur, future research will be necessary to more fully understand the effects of Medicaid expansion on health outcomes over time.
Economic Effects (Appendix Table C)
States expanding Medicaid under the ACA have realized budget savings, revenue gains, and overall economic growth. National-level and multi-state studies indicate that expansion states have experienced significant budget savings as individuals who were previously eligible for limited Medicaid benefits under pre-ACA eligibility categories transition to full Medicaid coverage in the new adult group, with enhanced federal funding.136,137 Furthermore, multiple expansion states have found that expansion increased state revenue from existing assessments on insurers and providers,138 and expansion states are realizing savings in areas of the state budget beyond Medicaid, such as state-funded behavioral health services and corrections.139,140
Single state studies demonstrate similar economic benefits from expansion. Some expansion states have documented increased general fund revenue141,142 and overall positive net budget impacts of expansion,143 as well as growth in state gross domestic product (GDP),144,145 real disposable personal income,146 and intermediate demand for goods and services.147 While these findings based on state expansion experiences to date reveal that Medicaid expansion has yielded a net fiscal benefit for states while the federal government has paid 100% of expansion costs, study projections suggest that annual state expansion costs will rise148,149 and at least one study shows annual deficits in later years as the state share of expansion costs reaches 10%.150 However, some projections show that despite the phase-in of the state share, states still expect to experience net revenue gains and economic growth into the future. Medicaid expansion in New Mexico is expected to generate a net surplus of over $300 million for the General Fund between fiscal year (FY) 2014 and FY 2021,151 Colorado’s economy is expected to be $8.53 billion (1.38%) larger as a result of Medicaid expansion by FY 2034-2035,152 expansion is estimated to have a positive cumulative impact of $30.1 billion on Kentucky’s economy through state fiscal year (SFY) 2021,153 and Arkansas is expected to see positive economic growth into the future regardless of how it funds its share of expansion costs beginning in 2017.154 An analysis conducted in Alaska also suggested that large scale opportunities exist to reduce state Medicaid costs over the long term to offset expansion costs.155
Despite Medicaid enrollment growth initially exceeding projections in many states, studies indicate that Medicaid spending per capita has been relatively low for the newly eligible expansion population and state general fund spending growth for Medicaid has been slower in expansion states compared to non-expansion states. As previously noted, Medicaid enrollment has exceeded projections in many states—one national study found that of the 29 states with expanded Medicaid programs in FY 2015, slightly more than half (17 states) noted that enrollment initially increased faster than expected.156 However, the same study found that nearly two-thirds of expansion states reported that per member per month costs for the expansion population were at or below projections, and the rate of growth in total Medicaid spending averaged 17.7% across the 29 expansion states, which was very similar to the 18.0% enrollment growth rate.157 Another national analysis found that spending per enrollee for the new adult eligibility group under the ACA Medicaid expansion was significantly lower than spending per enrollee across all Medicaid groups ($4,513 vs. $7,150).158 A separate study in Colorado found that while overall expansion costs exceeded projections in the state (primarily due to the unexpectedly high Medicaid enrollment following the expansion), per capita costs were lower than predicted. On average, each Colorado expansion enrollee cost approximately $4,100 annually in the first two years compared with the anticipated annual cost of $5,200.159 In addition, while not fully explained by the Medicaid expansion, growth in state general fund spending for Medicaid was slower in expansion states compared to non-expansion states (3.4% compared to 6.9% in FY 2015).160
State Medicaid expansions resulted in changes in the payer mix for hospitals and clinics. Studies using a range of data and approaches have documented steep declines in uninsured hospital visits161,162,163,164,165,166,167,168,169 and increases in Medicaid-covered visits following state Medicaid expansions.170,171,172,173,174,175,176,177 Coinciding with Medicaid expansion-induced coverage gains, studies consistently show reductions in the costs of uncompensated care in expansion states.178,179,180,181,182 In sharp contrast to their counterparts in expansion states, providers in non-expansion states appear to have experienced little or no decline in uninsured visits and uncompensated care during 2014 and subsequent years.183,184,185,186,187,188 Consequently, many providers in non-expansion states struggle with a range of financial challenges associated with the coverage gap and reductions in federal funding for uncompensated care.189,190
Additional evidence indicates that the financial benefits hospitals experience from expansion often outweigh increases in Medicaid shortfalls. A study of Ascension Health hospitals nationwide found that the decrease in charity care costs for hospitals in expansion states was greater than the increase in Medicaid shortfalls.191
While most studies in this area focus on hospitals, some evidence indicates that health clinics are also experiencing payer mix changes under expansion. Health clinics serving individuals who are homeless in expansion states have experienced increases in third-party payments as a share of total revenue, while similar clinics in non-expansion states remain heavily reliant on grant funding.192
Most studies show that the Medicaid expansion had a positive effect or no negative effects on employment and the labor market. Predictions of Medicaid expansion’s effect on employment and the labor market have varied, with some individuals and groups suggesting it will boost the economy and stimulate job growth and others arguing that it will have a negative effect on job growth and serve as a disincentive for low-income individuals to seek employment. Existing research on this topic is limited, however, and shows positive or neutral impacts of expansion. State-specific studies have documented193 or predicted194 significant job growth resulting from expansion. A study in Colorado found that the state supports 31,074 additional jobs due to Medicaid expansion as of FY 2015-2016.195 A study in Kentucky estimated that expansion would create over 40,000 jobs in the state through SFY 2021 with an average salary of $41,000.196 Additional studies examining other measures of employment and employee behavior (such as transitions from employment to nonemployment, the rate of job switches, transitions from full- to part-time employment, labor force participation, and usual hours worked per week) have not found significant effects of Medicaid expansion.197,198,199
Although states only began implementing ACA Medicaid expansions about two and a half years ago, a large body of literature analyzing the effects of expansion has developed. In general, this research suggests that expansion has had largely positive impacts on coverage, access to care and utilization, as well as economic outcomes, including impacts on state budgets, the payer mix for hospitals and clinics, and employment and the labor market. Yet, some studies do not find significant impacts and some point to challenges following expansion, such as provider shortages in some areas. These challenges may make it difficult to meet the increased demand for care among the newly-eligible Medicaid population.
Given that it is still early in the expansion implementation experience, ongoing study will be necessary to further assess and understand the full impact of expansion into the future. Expansion’s effects on health outcomes is one area in particular that calls for longer-term research. Additional research also will be important to assess the economic effects of state Medicaid expansions as states begin paying a share of expansion costs (the current 100% federal share will begin to phase down in 2017 until it reaches 90% in 2020) and as cuts in federal payments to states and providers for uncompensated care costs and Medicaid shortfalls go into effect. Despite the need for continued research in these and other areas, the beneficial effects of expansion documented in studies to date suggest that ACA Medicaid expansion presents a valuable opportunity for the 19 remaining non-expansion states to improve coverage, access to care and utilization, and a range of economic outcomes.Executive Summary Appendix