The Effects of the Medicaid Expansion on State Budgets: An Early Look in Select States
As enacted, the Affordable Care Act (ACA) broadened Medicaid’s role, making it the foundation of coverage for nearly all low-income Americans with incomes up to 138 percent of the federal poverty level (FPL) ($16,242 per year for an individual in 2015). However, the Supreme Court ruling on the ACA effectively made the decision to implement the Medicaid expansion an option for states. For states that expand Medicaid, the federal government will pay 100 percent of Medicaid costs of those newly eligible for Medicaid for up to three calendar years from 2014 to 2016. The federal share gradually phases down to 90 percent in 2020, where it remains well above traditional federal medical assistance percentage (FMAP) rates in every state. As of March 2015, 29 states (including the District of Columbia) have adopted the Medicaid expansion though debate continues in other states. A key issue for policy makers at the state level has been the state budget effects of the Medicaid expansion on states’ budgets.
This brief presents findings from a study of the early budget effects of the Medicaid expansion in three states: Connecticut, New Mexico, and Washington State. These interviews took place in the Fall of 2014, as executive-branch officials had begun preparing executive budget proposals for the 2015 legislative sessions. (See the Methodology section for more details on how the study was conducted.) Also included are findings from a separate study commissioned by Kentucky officials that examined the impact of that state’s decision to expand. The findings provide a limited and early insight into the effect of the Medicaid expansion on state budgets, both within and outside of the Medicaid programs. Key findings are summarized below, first looking at effects within the Medicaid budget, then turning to the effects on other parts of the state budget as well as revenues. A summary of the findings is also presented in the Summary Table. This study focused primarily on budget factors that may apply elsewhere, but one should be careful in generalizing, as each state’s budget situation is unique.
Medicaid Enrollment and State Costs
While enrollment among those previously eligible but not enrolled (which is financed at the state’s regular match rate) increased in each of the study states, the majority of this enrollment growth was driven by other changes in the ACA rather than just the Medicaid expansion. All states anticipated increased enrollment resulting from the Medicaid expansion, both for newly eligible adults and among those who were previously eligible but not enrolled. In each state, newly eligible enrollment exceeded expectations. Under the ACA, these costs are fully funded with federal dollars through December 31, 2016, so this did not increase state costs in SFYs 2014 or 2015. The extent to which states saw increased enrollment among those previously eligible but not enrolled varied across the study states. In Connecticut and New Mexico, the enrollment increase was above projections, but it was below projections in Washington State. A separate study commissioned by the Kentucky officials also found enrollment of those previously eligible but not enrolled was well above projections.1 However, the enrollment growth among those previously eligible but not enrolled in each of these states was primarily driven by other 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. Washington State officials, for example, estimated that nearly three-quarters of such enrollment growth resulted from features of the ACA that would have been present with or without expanded eligibility.
While study states saw some increased Medicaid administrative costs, these costs were relatively small. Medicaid administrative costs in general represent only five percent of total Medicaid spending nationally.2 Additionally, most of these administrative costs would have been incurred with or without the Medicaid expansion due to other aspects of the ACA. Officials expect these effects to be ameliorated by the ACA’s shift to a more data-driven and less labor-intensive approach to eligibility determination. The transition to this new approach is supported by 90 percent federal funding for necessary investments in information technology, along with 75 percent federal funding for operating expenses.3
Savings within Medicaid budgets
All study states reported savings within their Medicaid programs as some beneficiaries for whom states would have received standard FMAP instead qualified as low-income adults eligible for the ACA’s enhanced match rate.4 Conversion of limited Medicaid programs for low-income adults in each of the study states provided a source of immediate, significant savings within Medicaid programs. For two of these states, Connecticut and Washington, these limited Medicaid programs for low-income adults had started as state-funded indigent care programs that were converted to Medicaid financed programs (at the state’s regular matching rate) ahead of the ACA expansion. (More details on these programs are provided in Appendix A.)
In addition, some of the study states observed enrollment declines in optional Medicaid eligibility categories without reducing eligibility. For example, some study states saw declines in the enrollment of lower cost programs such as family planning (Connecticut and Washington) and breast and cervical cancer treatment programs (Washington). These two states also saw declining enrollment in higher-cost eligibility categories, such as medically needy spend-down programs for adults; Washington State also saw declining enrollment in an optional eligibility pathway that provides coverage for those awaiting an SSI disability determination. Adults who would have enrolled under these optional eligibility pathways were instead enrolling under the new Medicaid expansion group, qualifying for a higher matching rate.5 (See Appendix B for more on these pathways.) Similar declines in enrollment among optional groups were also seen in Kentucky; according to this separate report, the Commonwealth saw savings of over $38 million in SFYs 2014 and 2015 from beneficiaries qualifying under the newly eligible group instead of other optional pathways such as breast and cervical cancer treatment program and spend-down groups among others.6
Officials in Washington state also observed unexpected declines among pregnant women that had not been included in enacted budgets. Officials in Washington noted that much of the decline was due to more women qualifying under the Medicaid expansion group. Medicaid programs have long been required to cover pregnant women at levels at or above the Medicaid expansion. This requirement continues under the ACA; their coverage is reimbursed at the state’s regular match rate. However, women enrolled in the new adult expansion group who become pregnant are not required to move to the pregnancy-related eligibility group outside of their regular renewal period. Budget officials also noted that the availability or coverage in the Marketplace as well as improving economic conditions could also have caused some of this decline.
Savings Outside of Medicaid Budgets
All study states experienced savings in other areas of the state budget beyond Medicaid.7 Expanded Medicaid coverage helped to reduce some of the need for state-funded programs to serve this population, such as behavioral health and corrections. Savings either benefited the state general fund or were reinvested within the program area, often restoring cuts made during the economic downturn.
All three study states experienced savings in behavioral health programs. Behavioral health programs across the country saw substantial state funding cuts during the economic downturn; many remain underfunded.8 States that have implemented the Medicaid expansion may use the federal dollars from the Medicaid expansion either to substitute for state funds spent on mental health services, help restore funding cuts implemented during the economic downturn, or reduce general fund spending (e.g. “book” savings.) Connecticut and Washington State “booked” these savings for their general funds, while New Mexico reprogrammed the majority of savings within the behavioral health agency’s budget.
While the study states noted savings and efficiencies in their behavioral health programs due to the expansion, there were some challenges and delays in transitioning behavioral health care providers to billing for their clients’ claims (rather than relying on grant funding) and enrolling beneficiaries of behavioral health programs (a generally hard-to-reach population) into Medicaid. These challenges necessitated adjustments to original budget assumptions, but state officials were confident both that expansion was already yielding savings and that the magnitude of savings would likely grow as these transitions progressed.9 General Fund savings were also found for Kentucky as Medicaid beneficiaries – those newly eligible as well as those previously enrolled – received mental health treatment and substance use disorder services through community mental health centers reimbursed with Medicaid funds instead of general fund dollars.10 Coinciding with the Medicaid expansion, the Commonwealth of Kentucky had expanded the types of behavioral health providers that were eligible for Medicaid reimbursement, both for the traditional Medicaid program as well as for the Medicaid expansion, increasing access to such services.11
Two of the study states also experienced budget savings or offsets for corrections. Many inmates historically could not qualify for Medicaid since they did not fit into one of the traditional eligibility categories. Even for inmates who did meet the income and categorical eligibility requirements to qualify for Medicaid, federal law prohibits Medicaid payment for services provided in jails or prisons under a policy known as the “inmate exclusion.”12 However, Medicaid reimbursement is available for care provided to eligible individuals who are admitted to an inpatient facility off jail or prison grounds, such as a hospital, for at least 24 hours. Prior to the ACA, few states had pursued Medicaid reimbursement for these services given the limited share of the incarcerated population that could qualify for Medicaid.13 However, the Medicaid expansion offers greater potential savings to states from reimbursement for inpatient services provided to incarcerated individuals, since a larger share of the incarcerated population may qualify for Medicaid under the Medicaid expansion and the federal government is providing states an enhanced federal matching rate for newly eligible adults.14
Washington State included limited state budget savings in its enacted corrections budget for SFYs 2014 and 2015. Connecticut officials noted that while they hadn’t quantified savings from the Medicaid expansion, the state’s Department of Correction had been able to weather notable budget reductions in their inmate medical account since the state implemented the early expansion in 2010. New Mexico’s Medicaid program was working to realize savings in this area, but officials also noted that many corrections responsibilities are vested locally.15 General Fund savings from the Medicaid expansion in the state’s corrections department were also noted in a separate report commissioned by Kentucky.16
Some study states also reported savings in other areas, including uncompensated care payments and high risk pools. In addition to federal funding for uncompensated care costs through Medicare and Medicaid Disproportionate Share Hospital (DSH) programs, states and localities generally fund roughly 40 percent of uncompensated care costs.17 When the previously uninsured gain coverage that pays for their care, previously uncompensated costs decline. Among our three study states, only Connecticut had a state-level uncompensated care program in place before the ACA; Washington state did not have an uncompensated care pool and counties bear much of the responsibility for financing hospital uncompensated care in New Mexico. When it converted its pre-ACA state indigent care program into an early Medicaid expansion, Connecticut was able to significantly reduce their uncompensated care payments to hospitals as well as make some reductions in uncompensated care for community health and mental health centers. Reductions in state and local expenditures for uncompensated care were also noted through SFY 2016 in a separate report examining the expansion’s impacts on Kentucky; this same report also noted general fund savings in later years from the scheduled reductions in DSH funds.18 Early evidence from that state’s expansion also saw declines in uncompensated care charges as well as increased revenues for providers.19
Case study states noted additional areas of moderate or limited budget savings outside of Medicaid. For example, New Mexico, which was the only study state that operated a state-funded high-risk pool, saw moderate savings as enrollees transitioned to other coverage options. Like the movement of behavioral health program beneficiaries into Medicaid, this transition moved more slowly than expected, resulting in fewer short-term savings than originally projected. Connecticut and Washington also reported savings from state-funded public health programs; similar savings were also found in Kentucky as services provided through local health departments to Medicaid enrollees were now reimbursed by Medicaid.20 Washington reported savings from a state-funded program that provided long-term services for adults with developmental disabilities.
The impact on state revenue, as monitored by budget officials, was primarily reflected in increased provider and premium taxes and fees. Washington and New Mexico projected increased revenue from provider taxes and fees as a result of expansion in their budgets. Both states have premium taxes on insurers; revenues collected from these taxes and fees increased as more Medicaid members joined managed care plans and more Medicaid patients saw providers. Connecticut experienced no increase in revenues as a result of expansion. The state’s Medicaid enrollment has grown substantially as a result of the expansion and hospital revenues increased as care shifted from uncompensated to Medicaid-reimbursed. While Connecticut has Medicaid provider taxes and fees, they have not been rebased since 2009 and therefore have not increased due to Medicaid expansion. Further, as a state with very high rates of health insurance coverage prior to the expansion, insurance company premium taxes have not appreciably increased following expansion. As a final complicating factor, a state program that allows for the purchase of tax credits to reduce tax liability has been utilized by some providers subject to the Medicaid provider tax, and this has resulted in a reduction in revenue from this source.
Expansion is expected to increase overall economic activity,21 and thus state general revenue, due to the significant influx of federal Medicaid dollars used to purchase health care within such states. However, only one study state (New Mexico) specifically noted this effect in its overall economic and revenue projections for the first years of implementation. The other states’ economic forecasts did not include such detail in their underlying assumptions at this point. It will be difficult to isolate the expansion’s effects on work force and economic growth until more detailed data become available. However, a separate analysis conducted by the Urban Studies Institute at the University of Louisville estimated that the Medicaid expansion in Kentucky led to an increase of 12,000 jobs in SFY 2014 alone and over 40,000 additional jobs through 2021. The analysis estimates that this increase in jobs will result in additional tax revenue for the state an localities through SFY 2021.22
Disentangling the fiscal impact of expanded Medicaid eligibility from other ACA effects as well as other factors shaping health care costs can be a tremendous challenge. Policy and budget decisions are not made in a vacuum; isolating the budgetary effects of one policy decision from other policy decisions as well as from larger demographic and economic trends is inherently difficult. In this particular case, other changes resulting from the ACA, such as requirements that all states implement new policies to streamline and simplify Medicaid enrollment, the individual coverage requirement, and new coverage options available through the Marketplace make isolating the effects of the Medicaid expansion particularly difficult.
State budget offices are not set up to estimate the net budget impact of a single policy, such as the Medicaid expansion. Such offices, unlike the Congressional Budget Office, do not typically maintain alternative budget scenarios that estimate costs and revenues in the absence of a particular policy (such as the Medicaid expansion.) They rarely have good reason to spend resources analyzing the effects of past decisions, like expansion. Moreover, the fiscal effects of the Medicaid expansion are hard to analyze comprehensively because they are experienced across budget categories; cost implications fall within and outside Medicaid, and both general and special revenue sources can be affected. Among the three states examined, only Washington was in a position to assess the overall budgetary impact of expansion; at the time of our interviews, the state was in the final months of maintaining an alternative budget scenario that estimated state costs in the absence of Medicaid expansion. Based on that scenario, state savings from higher federal matching rates for newly eligible enrollees and from reduced spending on some (but not all) pre-ACA, state-funded programs could be analyzed. Without taking into account any revenue gains resulting from expansion, those savings exceeded increased state costs attributable to expansion in both SFY 2014 and 2015 (Appendix C); in fact, the state noted net Medicaid budget savings for each budget period throughout the 2013-2021, including the period during which the FMAP for low-income adults reaches its final 90 percent level. During SFY 2015, the net savings from expansion was projected to equal 1.7 percent of the state’s entire General Fund for SFY 2013.23 Kentucky commissioned Deloitte to examine the fiscal and economic impact of the Medicaid expansion decision on Kentucky; this independent analysis, which examined the impact of the expansion across the state’s budget and at the broader economic effects among other factors, estimates Kentucky will see a net positive fiscal impact of $919.1 million over the SFY 2014 through 2021 period compared to what the state would have spent in the absence of the Medicaid expansion.24
Early evidence from these case study states shows that expansion yields state savings and state revenues while causing limited increases in state costs. Both newly eligible consumers and those who qualified under pre-ACA categories can be expected to enroll in large numbers, although much of the latter enrollment will occur with or without expansion. States can experience notable savings both within Medicaid and outside Medicaid budgets, though savings in parts of the budget outside of Medicaid may be slower to materialize than anticipated, and policymakers may choose to reinvest savings to increase the provision of non-Medicaid services rather than reduce General Fund commitments. Two of the study states projected increased revenue from provider taxes and fees; states are expected to also realize revenue gains from increased economic activity as evidenced by the findings of the expansion’s impact on Kentucky. In sum, our analysis of early experiences in three states suggests that expansion creates both state budget savings and some limited initial costs for states in these early years of the expansion, when the cost of the newly eligible is fully financed with federal dollars.
This brief provides insight into the early experiences in only three states along with findings from a separate study commissioned by Kentucky. Each state and its budget are unique. The findings of this brief are likely to illustrate important general trends, but ultimately the effect of Medicaid expansion on state budgets must be assessed in terms of the particular circumstances of each state. In states that have already chosen to expand eligibility, the implications of that decision on state budgets and revenues will continue to be monitored as implementation continues and more data become available.
This brief was prepared by Stan Dorn and Norton Francis of the Urban Institute and Robin Rudowitz and Laura Snyder from the Kaiser Family Foundation.
The authors also wish to thank the state budget officials and staff in Connecticut, New Mexico and Washington State who participated in this study. Especially in this time of limited resources and challenging workloads, we truly appreciate the time and effort provided by these public servants to participate in structured interviews and respond to our follow-up questions. Without their generous assistance, this brief would not have been possible.
This study analyzes the state budgetary effects that have been identified thus far in three geographically diverse states that began implementing the full expansion on January 1, 2014: Connecticut, New Mexico, and Washington State. Researchers from the Kaiser Commission on Medicaid and the Uninsured and the Urban Institute interviewed state budget staff and officials and reviewed state budget documents during August through November, 2014, before the start of the 2015 open enrollment period. These interviews were based on semi-structured protocols, and key topic areas were shared in advance with state officials.
The interviews took place as executive-branch officials had begun preparing executive budget proposals for forthcoming legislative sessions. Of the three states included in this study, two (Connecticut and Washington state) operate on a biennial budget cycle while New Mexico operates on an annual budget cycle. All three states were in the process of developing new budgets for the next budget window, which would cover SFY 2016 as well as SFY 2017 in Connecticut and Washington. All three study states adopted the Medicaid expansion as of January 1, 2014, halfway through their SFY 2014 budgets. Because of the timing of this study, states were asked about the budget effects for SFY 2014 and SFY 2015. While SFY 2014 had ended, states were still finalizing actual figures for SFY 2014. Therefore, state officials were basing responses off of what was included in SFY 2014 and SFY 2015 budgets as enacted; for SFY 2014, officials commented where they could on what had been observed (e.g. if savings originally included in budgets were in line with original assumptions.)
Washington State was able to provide cost and savings estimates for a number areas affected by the state’s decision to implement the Medicaid expansion due to the fact the state has maintained an alternative budget scenario that estimated state costs in the absence of Medicaid expansion, a process the state is expected to stop in the near future. Budget officials in Connecticut and New Mexico reported cost and savings estimates where possible, but all states reported more broadly about the scope of changes that had been considered to date. Each of these states reviewed the findings; their feedback has been incorporated.
Additional findings of the expansion’s impact on Kentucky published in a separate report commissioned by that state have also been included.
This study focused primarily on budget factors that may apply elsewhere, but one should be careful in generalizing, as each state’s budget situation is unique.