ABSTRACT
- Issue: The impact of Medicaid expansion on state budgets is a concern cited by policymakers in nonexpansion states.
- Goal: To estimate the financial impact of Medicaid expansion on state budgets based on state experiences to date.
- Methods: Using historical data, projections of cost, and difference-in-differences analysis, we estimate the impact of expanding eligibility for Medicaid on states’ spending on the program and the overall effect on their budgets.
- Key Findings: During 2014–17, Medicaid expansion was associated with a 4.4 percent to 4.7 percent reduction in state spending on traditional Medicaid. Estimates of savings outside of the Medicaid program vary significantly. Savings on mental health care, in the corrections system, and from reductions in uncompensated care range from 14 percent of the cost of expansion in Kentucky to 30 percent in Arkansas.
- Conclusion: It is not necessary to cut other spending or raise revenue by 10 percent of the cost of expansion — their share in 2020 — to balance their budgets. States have a variety of means to offset some or all of expansion’s statutory costs. Thus, the net cost of Medicaid expansion to states is different from the “sticker price.” In some cases, the net cost is negative.
Introduction: How Does Medicaid Expansion Affect State Budgets?
Is expanding eligibility for Medicaid a good deal for states? This question has loomed over state policymakers for more than a decade. While hundreds of studies have detailed the costs and benefits of Medicaid expansion in terms of access to care, fiscal impacts, health outcomes, and other factors, consensus remains elusive.1 Many skeptics still question whether expansion is worthwhile from a budget perspective. Fourteen states still have not expanded Medicaid, and support for expansion in states that have done so is not unanimous.
In this brief, we focus on a common barrier to Medicaid expansion: concerns about its impact on state budgets. Expansion opponents often argue that its fiscal cost is too high. For the first three years, the federal government paid the full cost of expansion. States began covering a portion of expansion’s cost in 2017 and, starting in 2020, are responsible for 10 percent of its cost. Given recent spending levels, expansion states will collectively pay more than $7 billion in 2020. For the median expansion state, expansion will cost more than $100 million.2
These costs represent the “sticker price” of expansion. However, its actual fiscal impact differs from the sticker price for three reasons. First, expanding eligibility allows states to cut spending in other parts of their Medicaid programs. Second, it allows states to cut spending outside of Medicaid — particularly on state-funded health services for the uninsured. Finally, expansion may increase state revenues due to taxes related to Medicaid expansion or taxes on the increased economic activity it triggers.
To date, dozens of studies have documented the fiscal effects in expansion states. While the studies do not account for every possible impact, all find that the net cost of Medicaid expansion is well below the sticker price. In many cases, researchers have found that Medicaid expansion generates enough savings and/or new revenue to more than offset a state’s share of the cost. Building from these studies, researchers in at least eight nonexpansion states project similar savings and revenue, should their states expand Medicaid.3
In this brief, we summarize prior research documenting the impacts of Medicaid expansion on state budgets, focusing on a handful of states that have recent projections for expansion’s expected costs in 2020 and beyond, when states are responsible for 10 percent of the costs of expansion. We also present the results from new analyses of Medicaid expansion’s fiscal impacts based on budget data from all 50 states.
Ultimately, while each state’s Medicaid program is unique and the impacts of expansion vary from state to state, the accumulated information points in a single direction: states do not pay the full cost of Medicaid expansion. The net impact on their general funds is much smaller than expansion’s costs. In some cases, Medicaid expansion more than pays for itself.
The Impact of Medicaid Expansion on State Spending
States must finance a share of the cost of expansion. As such, expanding Medicaid will increase state spending. However, expanding Medicaid also allows states to reduce spending on traditional Medicaid. Thus, the net increase in total Medicaid spending is smaller than the cost of expansion. Expanding Medicaid also may allow states to cut spending outside Medicaid — particularly on programs that provide health services to low-income people.
Changes in State Medicaid Spending
Exhibit 1 illustrates the cumulative effects of expansion on states’ Medicaid spending using data from the National Association of State Budget Officers (NASBO).4 Excluding the first state fiscal year of expansion (which was half preexpansion and half postexpansion), Medicaid expansion increases total Medicaid spending by approximately 23 percent and federal Medicaid spending by approximately 38 percent, but it does not increase state Medicaid spending (at least not through the first five-and-a-half years).
The teal bars show the two competing effects of expansion on state Medicaid spending. During FY2015 and FY2016, the federal government paid the full cost of expansion and Medicaid spending in expansion states declined by approximately 6 percent relative to nonexpansion states. In FY2017, states began paying a share of the cost of expansion (5 percent), which grew each year between FY2017 and FY2020. As the share of Medicaid expansion costs paid by the states grew, net savings fell. However, at least through FY2019 in the average state, expansion generated sufficient savings to states’ traditional Medicaid programs to offset its costs.
This suggests that Medicaid expansion generates substantial savings in states’ traditional Medicaid programs because states’ expansion costs are significant. As shown in Exhibit 2, state spending on Medicaid expansion exceeded $4 billion in 2018 and, if spending remains at current levels, will exceed $7 billion in 2020 (among only the states that had expanded by 2018).
How does Medicaid expansion generate savings to offset such costs? Medicaid expansion allows states to access an enhanced federal match for some people who would otherwise be covered by traditional Medicaid (Appendix B). Given that states pay between 25 percent and 50 percent of the cost for a traditional Medicaid beneficiary but only 10 percent of the cost for an expansion beneficiary, these savings can be substantial. States can save from 15 cents to 40 cents on every dollar of care it can shift to expansion (assuming 2020 expansion match rates).
Prior research identifies several types of expansion beneficiaries who would likely receive traditional Medicaid in the absence of expansion.5 Some are people who would have been covered through specific Medicaid programs, such as Section 1115 waivers or the Breast and Cervical Cancer Treatment Program. Some are people who used to spend down assets or pursue a disability designation to qualify for Medicaid that now qualify for expansion. The populations affected vary from state to state, depending on the specifics of their traditional Medicaid programs.
While researchers have a clear understanding of how expansion may reduce states’ traditional Medicaid spending, calculating the magnitude of these savings is difficult. In expansion states, we only observe what happened with expansion — not what these states would have spent had they not expanded Medicaid.
One common approach to estimating the effects of Medicaid expansion uses nonexpansion states as a comparison group. If one assumes that, in the absence of expansion, traditional Medicaid spending in expansion states would have followed the same trajectory as spending in nonexpansion states, then one can estimate the effects of Medicaid expansion by subtracting the change in nonexpansion states from the change in expansion states. This is a difference-in-differences approach, and it has been used to identify the effects of Medicaid expansion on a variety of outcomes.6 In Exhibit 3, we present results from a difference-in-differences analysis that examines the effect of expansion on traditional Medicaid spending (i.e., total Medicaid spending minus total expansion spending), using data through federal fiscal year 2017 (see Appendix A for methods).
We present results from two different specifications in the table. The first column uses all expansion and nonexpansion states. The second column excludes states that expanded during 2015–17. Excluding these states avoids problems associated with control states becoming intervention states. It also avoids the potential to underestimate effects due to the transition period that occurs after expansion but before the full savings are realized. Both specifications yield similar results. During 2014–17, Medicaid expansion was associated with a 4.4 percent to 4.7 percent reduction in state spending on traditional Medicaid.
Other analyses using different methods to estimate savings within Medicaid find similar effects. Some estimate savings by assuming that spending levels for certain programs or populations would have remained at preexpansion levels.7 Others estimate savings by trying to identify the number of expansion beneficiaries who would have qualified for traditional Medicaid in the absence of expansion and then assume expansion saves states a portion of the cost of caring for these people.8 Both approaches have limitations. It is possible that spending would not have remained the same in the absence of expansion. It is also difficult to identify all those who may have enrolled in traditional Medicaid in the absence of expansion.
Exhibit 4 presents several estimates based on these approaches. While effects vary across states, studies of this type find that Medicaid expansion generates savings that average to 4 percent to 5 percent of traditional Medicaid spending, similar what we found in our difference-in-differences analysis.
The final column illustrates how much of states’ expected expansion costs in 2020 might be offset by savings within Medicaid. These estimates are crude because they are based on incomplete data (e.g., some states include some types of potential savings, but not others). We also do not have official forecasts for expansion spending in each state in 2020, so some estimates are derived from recent expansion spending. The results in this table suggest that savings to traditional Medicaid offset much of the expected cost of Medicaid expansion. In some cases, they offset nearly all the costs.
As the state share increases in 2020, savings within Medicaid will shrink. Increasing the state share reduces how much states save by covering someone in expansion who would otherwise receive traditional Medicaid. It also increases the state cost for Medicaid expansion. However, these data suggest that, on average, much of the cost of Medicaid expansion will continue to be offset by other savings in the program.
Savings Outside of Medicaid
When states expand Medicaid, they may see reduced spending outside of the program. Many states provide health care services to low-income residents; expansion may allow them to provide some of these services via Medicaid.
Prior studies identify several areas where expanding Medicaid reduces other state spending.9 The three most common include:
- Mental health and substance abuse treatment: Many states directly support mental health and substance abuse treatment for low-income people without health insurance. With Medicaid expansion, recipients may obtain these services via Medicaid.
- Corrections: Medicaid expansion allows states to shift the cost of some inmates’ health care from the state corrections budget to Medicaid.10
- Uncompensated care: Many states help offset the cost of providing care to people who cannot pay their medical bills. By reducing the number of people without insurance, Medicaid expansion significantly reduces the amount of uncompensated care.11 Therefore, some states have chosen to reduce payments to health care providers for uncompensated care.
This list is not exhaustive. Individual states report savings in a variety of other programs.
Any attempt at quantifying non-Medicaid budget savings due to expansion faces several hurdles. First, one needs to identify which budget changes are attributable to Medicaid expansion. This is sometimes straightforward; when a state expands Medicaid and then eliminates a mental health program for low-income, uninsured people, the ensuing savings are attributable to expansion. However, the connections are often less clear. Also, to the extent that Medicaid expansion changes outcomes (e.g., in terms of crime or employment rates), it may change the level or type of services states choose to provide. Identifying these changes and attributing them to Medicaid expansion among all the other moving pieces of state budgets are difficult. As such, estimates of savings are likely to be imprecise.
Despite these limitations, several studies document substantial savings to state budgets outside of Medicaid. Exhibit 5 includes estimates for several categories of non-Medicaid savings.
As illustrated in Exhibit 5, estimated savings vary considerably among states. However, while the savings associated with individual categories may be small, they often add up to a substantial amount. In Kentucky, these categories of savings offset roughly 14 percent of Medicaid expansion’s costs. In Arkansas they offset 30 percent and in Michigan they offset 41 percent.12
Exhibit 5 presents estimates from just three categories of potential savings. Expansion also may yield savings in areas including public health or indigent care, and such savings could be large. For instance, California saved $750 million in 2015 by reducing funding for a program that provided care to medically indigent adults who did not qualify for Medicaid.13 Pennsylvania anticipated savings of $626 million in FY2016 from cuts to general assistance medical coverage for people not eligible for Medicaid.14
Where possible, statistical analyses like those conducted above find effects similar to those reported in Exhibit 5. For instance, Exhibit 6 presents data on state corrections’ department health care spending through FY2015. These data show significant savings in Medicaid expansion states relative to nonexpansion states.15 Between FY2013 and FY2015, spending on health care for people in the corrections system among states that had not expanded by FY2015 increased by an average of 10 percent. However, in states that expanded in 2014, such spending increased by an average of 4 percent. Assuming that expansion states would have followed the same trajectory as nonexpansion states, this suggests that Medicaid expansion reduced state corrections health spending by approximately 6 percent. Shifting 6 percent of corrections health care spending to Medicaid and assuming the state pays for 10 percent of these costs, the net reduction in state spending amounts to roughly 5 percent of the expected costs of Medicaid expansion. This amount is similar to the savings reported in Exhibit 5, suggesting that the savings reported there are fairly typical.
Looking broadly at total non-Medicaid spending among states using the NASBO data and methods used in Exhibit 1, we do not observe significant effects associated with Medicaid expansion. In some years, Medicaid expansion is associated with an increase in total non-Medicaid spending. In other years, it is associated with a decrease in total non-Medicaid spending. In all years, the effect is not precise, so the results are not statistically significant.
It is not surprising that we do not observe statistically significant declines in states’ non-Medicaid spending associated with expansion, since other contemporaneous events obscure these savings. First, some states choose to leverage Medicaid expansion to increase total revenue, as discussed below. In these states, total non-Medicaid expansion spending might increase as states fund new priorities. Second, states that realize savings outside Medicaid may choose not to reduce total spending and instead transfer savings to other parts of the state budget. Finally, given that the savings outside Medicaid amount to a fraction of a percent of non-Medicaid spending, it would require a much larger data set to find statistically significant results.
While it is difficult to establish a precise estimate of the savings expansion creates outside of Medicaid, ample evidence suggests it does allow states to reduce spending in some non-Medicaid areas and offset some of the cost of expansion.
Montana: Checking All the Boxes
Montana has experienced significant savings in its traditional Medicaid program and modest savings outside Medicaid related to expansion. To help pay for it, the state has imposed provider taxes and premiums. Research also documents significant economic impacts and associated tax revenues tied to expansion.
Eliminating Medicaid expansion in Montana would increase spending and reduce revenues by an amount approximately equal to 123 percent of the state’s expected share of Medicaid expansion in FY2021. While actual impacts and costs will likely deviate from these estimates, Montana has a variety of pathways to pay the full statutory cost of Medicaid expansion without needing to cut programs or raise additional revenues.
Summary of Fiscal Impacts of Medicaid Expansion in Montana Costs, savings, or revenues ($ millions) % of expected state expansion costs Expected costs (benefits and administration) $66.67 Savings to traditional Medicaid $28.5 43% Savings outside traditional Medicaid $8.7 13% Hospital tax $15 22% Taxes on increased economic activity $25.2 38% Premiums $4.6 7% Total savings + revenues $69.8 123% Notes: All savings are estimates except hospital tax obtained from Ward and Bridge (2019). Hospital tax estimates obtained from MTN News, “Hospitals Pay Tax in Medicaid-Expansion Bill But Make Millions More Through Other Means,” KPAX, Mar. 28, 2019; updated Mar. 29, 2019. Expected state costs are for state FY 2021. We note that these estimates are crude. Multiple parties have developed estimates for some of the impacts included here, and their estimates do not always precisely align (although they are usually similar).
Revenue Effects
Medicaid expansion also affects the revenue side of states’ budgets. It may boost revenues in three ways: 1) states may impose expansion taxes or may have provider taxes that grow naturally with expansion; 2) if Medicaid expansion impacts the larger economy (e.g., resulting in more jobs), these impacts will generate more revenue; and 3) some states push some of the cost of expansion onto beneficiaries by charging premiums.
Taxes or Fees Tied to Medicaid Expansion
Several states have explicitly raised taxes and fees to cover their share of Medicaid expansion. Other states already had provider taxes or fees that grew naturally with expansion. According to the most recent 50-state Medicaid Budget Survey, 11 states fund the state portion of expansion with new or expanded taxes or fees.16 However, nearly every state has at least one type of provider fee used to pay for Medicaid, and several have expanded or changed these taxes/fees since Medicaid expansion (such as California, Oregon, and Illinois).17 These taxes or fees are often explicitly tied to Medicaid expansion. However, they are also used to fund traditional Medicaid, and state funding data often do not allocate funds between different parts of Medicaid. This makes it difficult to quantify the share of these taxes that directly fund expansion.
Exhibit 7 describes several taxes implemented by states to cover all or part of the state share of Medicaid expansion. The table is limited to taxes/fees where the amount of revenue allocated to expansion is clearly stated. Consistent with the 50-State Medicaid Budget Survey, several states impose taxes that raise sufficient revenue to pay the full sticker price of expansion.
It is important to note that these taxes do not suggest that states do not also reap the savings described above. Some states have chosen to leverage Medicaid expansion to raise revenue that allows them to fund other priorities or reduce other taxes. In fact, a recent income tax cut in Arkansas was linked to budget savings created by Medicaid expansion.18
Michigan: Full Offset Required
Michigan expanded Medicaid under the Social Welfare Act, which included a provision that automatically sunsets the program whenever annual state savings are insufficient to cover the costs. As a result, Michigan closely tracks the fiscal impacts of expansion. In its most recent estimates, Michigan’s House Fiscal Agency estimated that it would generate sufficient savings to offset expected costs through at least 2027–28 (the final year examined). The estimated net savings generated by expansion are substantial, amounting to more than $160 million per year.19
Michigan’s official estimates likely understate the savings within traditional Medicaid. The House Fiscal Agency includes the elimination of its Adult Benefits Waiver, but it does not include any other potential savings within Medicaid as described above.
Michigan’s official estimates also do not account for increased revenues attributable to increased economic activity. Levy et al. (2020) argue that Michigan likely enjoys nearly $140 million per year in additional tax revenue because of expansion. While some of this revenue may be offset by the costs associated with providing services to a larger economy, the net effect is still large.
Summary of Fiscal Impacts of Medicaid Expansion in Michigan Costs, savings, or revenues ($ millions) % of expected state expansion costs Expected costs (benefits and administration) $456 Savings to traditional Medicaid $47 10% Savings outside traditional Medicaid $188 41% Insurance provider assessment $171 38% Other restricted revenues $230 50% Taxes on increased economic activity $135.9 30% Total savings + revenues $771.9 169% Data: All estimates for FY2021 from Koorstra (2018) except taxes on increased economic activity, which comes from Levy et al. (2020).
Revenue Impacts from Increased Economic Activity
States that decide to expand Medicaid turn on a spigot of federal funds. States that do not expand Medicaid do not receive a special tax break or grant equal to the amount of federal Medicaid dollars they have forgone. Thus, at the margin, the decision to expand Medicaid is in part a decision to bring a substantial amount of money (and the associated economic activity) into the state’s economy.
When new money enters a state’s economy, it boosts employment and income. Several studies have calculated the impact of Medicaid expansion on states’ economies.20 Most employ a statistical model used to estimate the number of jobs and amount of income generated by an event like Medicaid expansion. Using this method, researchers have estimated the economic impacts of Medicaid expansion in Michigan, Montana, Louisiana, Colorado, Kentucky, New Mexico, and Arkansas.21 These studies consistently find that Medicaid expansion generates significant economic impacts.
One national study uses a difference-in-differences approach to estimate the effect of Medicaid expansion on total employment in states where expansion had a large effect on overall insurance coverage.22 This study finds similar results as do the economic impact analyses. Medicaid expansion leads to significant increases in the size of expansion states’ health care sector (approximately 3%) and total employment (1.3% in the fourth year after expansion).
More economic activity generally yields more state revenue.23 Recent estimates from Louisiana, Michigan, and Montana suggest that the economic impacts of Medicaid are sufficient to generate tax revenues equal to 30 percent to 37 percent of states’ expansion costs.24 These same forces also boost local tax revenues. Richardson et al.’s 2019 study of Medicaid expansion in Louisiana found that it maintained or supported $60.6 million in local tax receipts in FY2018.
Virginia: Paid for Twice Over
Virginia did not expand Medicaid until 2019. As a late-expanding state, it could learn from earlier states when developing estimates for the likely fiscal impacts of expansion. Prior to expanding Medicaid, Virginia expected to save nearly $270 million in FY2020. However, after one year of expansion, cost savings attributable to Medicaid expansion were even larger than expected. For FY2020, the governor’s amended budget includes additional savings of $211.7 million.25 This additional savings is predominantly attributed to unexpectedly large savings from people switching from traditional Medicaid (with a 50 percent state share) to expansion (with a 10 percent state share).
Even though Virginia expected (and has realized) substantial savings associated with Medicaid expansion, it still implemented a provider fee to cover its full statutory cost. Accordingly, between new revenues and savings, Virginia has likely “paid” for the cost of expansion twice over without including any increased revenue attributable to more economic activity.
Summary of Fiscal Impacts of Medicaid Expansion in Virginia Costs, savings, or revenues ($ millions) % of expected state expansion costs Expected costs (benefits only) $307.2 Savings to traditional Medicaid $221.4 (initial)
$433.1 (revised)72% (initial)
141% (revised)Savings outside traditional Medicaid $51.9 17% Insurance provider assessment $318.6 104% Taxes on increased economic activity N/A Total savings + revenues $591.9 (initial)
$803.6 (revised)193% (initial)
262% (revised)Data: Virginia Department of Medical Assistance Services, Overview of the Governor’s Introduced Budget: Presentation to Senate Finance Committee Subcommittee on Health and Human Resources (State of Virginia, Jan. 8, 2018); and Senate Finance & Appropriations and House Appropriations Committees (State of Virginia, 2020).
Expansion Premiums
Finally, states can generate revenue to offset expansion costs by pushing some costs onto beneficiaries via premiums or cost-sharing.26 As of 2019, five states charged premiums to beneficiaries covered through expansion and 25 imposed some form of cost-sharing.27 The federal government limits how much states can charge particular populations. As such, the effects of premiums and cost-sharing are limited. In some states, premiums offset approximately 10 percent of state expansion costs.28 However, some research suggests that premiums deter enrollment and increase state administration costs.29 Therefore, full consideration of the impact of premiums on the fiscal cost of Medicaid expansion should also account for spillover effects such as administrative costs, a smaller Medicaid expansion program, lower potential savings to traditional Medicaid or other programs, and potentially smaller economic impacts.
Ohio: Offsetting Some, But Not All, Costs
Ohio’s experience differs to some degree. Its official savings estimates amount to only 24 percent of the cost of expansion in FY2021. While this estimate does not fully capture every dollar of savings, it does suggest that savings may be lower in some states.
Ohio funds Medicaid expansion with revenues from two health care taxes tied to expansion. These taxes are expected to generate $248 million in FY2021, or 46 percent of the state’s share of expansion costs. Combined, these factors offset 70 percent of the cost of expansion. This means that, as of FY2021, Ohio expects to pay 3 percent of the total cost of expansion. While uncounted savings and/or tax revenues from economic impacts may shrink the gap, the Ohio experience suggests that some states may face a marginal cost from expanding Medicaid.
Summary of Fiscal Impacts of Medicaid Expansion in Ohio Costs, savings, or revenues ($ millions) % of expected state expansion costs Expected costs (benefits only) $534 Savings to traditional Medicaid $36 7% Savings outside traditional Medicaid $90 17% MCO taxes $248 46% Taxes on increased economic activity N/A Total savings + revenues $374 70% Data: Ohio Office of Budget and Management analysis, July 2018.
Conclusion
Studies that examine the fiscal impact of Medicaid expansion on specific states or the effects across all states find consistent results: expansion leads to significant budget savings and significant revenue increases (even without imposing additional taxes). Consequently, the actual net price of expansion is well below the sticker price to states of 10 percent. In some cases, states’ net price is negative. Medicaid expansion can provide states with additional resources to fund other priorities or cut taxes.
Given that each states’ Medicaid program is different, fiscal effects and pathways to savings vary widely. Not every state that expands Medicaid will experience large savings in their traditional Medicaid programs, but many will. Not every state will experience large savings outside Medicaid, nor revenue growth, but some will. While the paths may differ, the available evidence points in the same direction: states pay, at most, only a small fraction of the cost of Medicaid expansion.
ACKNOWLEDGMENTS
The author would like to acknowledge Akeiisa Coleman, Rachel Nuzum, and Audrey McIntosh at the Commonwealth Fund for their helpful comments and suggestions.
