As the debate over amending health insurance market rules continues, proponents of changing the law have proposed reducing the health benefits provided by non-group plans as a potential way to lower premiums in the market.  The Affordable Care Act (ACA) prescribes 10 categories of essential health benefits that non-group and small-group policies must cover, and provides in most cases that the scope of these benefits should be similar to those in employer group health plans, which cover most non-elderly Americans.  The American Health Care Act (AHCA), which passed the House of Representatives on May 5, would permit states to seek waivers to amend the required benefits if doing so would achieve one of several purposes, including lowering premiums.¹   We look below at the benefits covered by non-group plans before the ACA as a possible indication of how states could respond to the waiver authority under the AHCA.

Background

The lack of coverage for benefits such as maternity and mental health care in many nongroup plans, which was a frequent point of criticism when the ACA was debated, was one (but not the only) reason why non-group coverage was less expensive before the ACA was enacted.  In the pre-ACA market, certain benefits were excluded to make coverage more affordable and to guard against potential adverse selection by applicants with more predictable, chronic health care needs.  Even with the ability to medically screen applicants for non-group policies, some insurers excluded coverage for conditions such as mental health and substance abuse care unless states required that they be covered.

States determined coverage requirements for health insurance policies prior to the ACA.  A few states defined a standard benefit package to be offered by insurers in the nongroup market.  Most states adopted some mandates to cover or offer specific benefits or benefit categories – such as requirements for policies to cover maternity benefits or mental health treatments. In addition to deciding which categories of benefits must be included or offered, states might also specify a minimum level or scope of coverage; for example, a few states required that mental health benefits have similar cost sharing and limits as other outpatient services (sometimes called parity).

Pre-ACA non-group plans varied considerably in scope and comprehensiveness of coverage, with some plans limiting benefit categories or putting caps on benefits, while others offered more comprehensive options.  For example, some plans did not cover prescriptions, others covered only generic medications or covered a broader range of medications subject to an annual cap, while still others covered a more complete range of medications.  This diversity was possible because insurers generally were able to decline applicants with pre-existing conditions, and could require their existing customers to pass screening if they wanted to upgrade to more comprehensive benefits.  This prevented applicants from selecting the level of coverage they wanted based on their known health conditions, but also prevented many people from being able to obtain non-group coverage at all.

To look more closely at the benefits provided in pre-ACA non-group plans, we analyzed data submitted by insurers for display on HealthCare.gov for the last quarter of 2013.  Beginning in 2010, insurers submitted information about their non-group plans to be displayed on HealthCare.gov; the data includes information on benefits, coverage levels for each benefit, benefit limits, premiums and cost sharing parameters, and enrollment.  We focus here on the benefits and benefit limits.  We use data from 2013 because it is the most current year prior to when the ACA’s major insurance market changes went into effect, provides more benefit categories than some earlier years, and has more information about benefit limits for each category.  We note, however, that the ACA prohibition on annual dollar limits took effect shortly after enactment and was phased in between 2010 and 2013, so these types of limits would likely not be reflected often in data we received. This means that our analysis likely misses some of the limits (for example, dollar limits on prescriptions) that existed in nongroup policies before the ACA was enacted.  We limit the analysis to plans where insurers report enrollment in the product upon which the plan is based.  Our methods are described in more detail in the appendix.

Results

The data include 8,343 unique plans across 50 states and the District of Columbia.  We looked at the percentage of plans that included coverage for major benefit categories.  Not surprisingly, all of the plans covered basic benefits such as inpatient hospital services, inpatient physician and surgical services, emergency room services, and imaging services, while virtually all (99%) covered outpatient physician/surgical services,  primary care visits, home health care services, and inpatient and outpatient rehabilitation services.

Certain other benefits, however, were covered much less often (Figure 1).  Large shares of plans did not provide coverage for inpatient or outpatient mental/behavioral health care services (38% each), inpatient or outpatient substance abuse disorder services (45% each), and delivery and inpatient care for maternity care (75%).²  In addition, 6% of plans did not provide coverage for generic drugs, 11% did not provide coverage for preferred brand drugs, 17% did not provide coverage for non-preferred brand drugs, and 13% did not provide coverage for specialty drugs.

Even when coverage was provided, some policies had meaningful limits or restrictions for certain benefits.  Mental/behavioral health care is a case in point.  Among plans with coverage for outpatient mental/behavioral health services, 23% limited benefits for some or all mental/behavioral services to fewer than 30 visits or sessions over a defined period (often a year) and 12% limited it to 12 or fewer.  A small share (about 5%) of plans providing coverage for outpatient mental/behavioral health services provided benefits only for conditions defined as severe mental disorders or biologically-based illnesses or applied limits (such as visit limits) if the illness was not defined as severe or biologically based.  The definitions of these terms varied by state.³ 

Similarly, for plans covering outpatient substance abuse disorder services, 22% limited the benefit to fewer than 30 visits or sessions; 12% limited it to 12 or fewer. In many of these plans, visits for either mental health or substance abuse care were combined to apply toward the same limit.

Among the relatively few plans that provided coverage for delivery and inpatient maternity care, a small share (3%) applied separate deductibles of at least $5,000 for maternity services and some plans (6%) applied a separate waiting period of at least year before benefits were available.  A few plans restricted benefits to enrollees enrolled in family coverage or required that the enrollee’s spouse also be enrolled.

Discussion

The ACA raised the range of benefits provided by non-group policies such that the benefits now offered by non-group plans are comparable to those offered in employer group plans.  The desire to lower non-group premiums, however, has led policymakers to consider allowing states to roll back the essential health benefits prescribed by the ACA.

Among the pre-ACA policies we reviewed, virtually all included benefits for certain services: hospital, physician, surgical, emergencies, imaging, and rehabilitation.  Other services were covered less often, including prescription drugs, mental/behavioral health care, substance abuse disorder care, and coverage for pregnancy and delivery.  This latter group of services all have some element of predictability or persistency that make them more subject to adverse selection. For example, many people use drug therapies over long periods and would be much more likely to select policies covering prescriptions than people who do not regularly use prescription drugs. If states were to drop any of these services from the list of essential health benefits for non-group plans, access to them could be significantly reduced.

The difficulty is that insurers would be very reluctant to offer some of these services unless they were required in all policies because people who need these benefits would disproportionately select policies covering them. In the pre-ACA market, insurers were able to offer products with different levels of benefits because they generally were able to control who could purchase them by medically screening new applicants.  Even existing customers faced medical screening if they wanted to change to a more comprehensive policy at renewal.  Through these practices, insurers were able to avoid the situation where people could choose cheaper policies when they were healthy and upgrade to better benefits when their health worsened. The proposed AHCA market rules, however, would not guard against this type of adverse selection, because people with pre-existing health conditions would be able to select any policy offered at a standard premium rate, and change their selection annually without incurring a penalty, as long as they maintained continuous coverage. This means that the range of benefits provided by insurers in states with essential health benefit waivers would likely be more limited than what insurers offered in the pre-ACA non-group market.  Benefit choice might be particularly limited in states that specify only a few benefits as essential.

It is hard to imagine that insurers would cover certain benefits if they were not required.  For example, some insurers before the ACA did not offer mental health benefits unless required by a state, even when they could medically screen all of the applicants.  And given the current problems with substance abuse in many communities, insurers would be reluctant to include coverage to treat them unless required. Offering these benefits as an option (for example, including them in some policies but not in others), would result in very high premiums for optional benefits because people who know they need them would be much more likely to choose them.

The AHCA presents state policymakers with a dilemma: they can reduce the essential health benefits to allow less expensive insurance options for their residents, but doing so may eliminate access to certain benefits for people who want and need them.

The data we used in our analysis were obtained from the Centers for Medicare and Medicaid Services (CMS) through a Freedom of Information Act Request, submitted January 10, 2017, with data supplied April 11, 2017.  The data are now available here: https://www.cms.gov/cciio/resources/data-resources/health_plan_finder_data.html. Insurers submitted the data to CMS to be displayed on the Health Plan Finder on HealthCare.gov.  Data were submitted periodically and updated periodically by CMS. We received data for three quarters in 2011, and for four quarters in 2012 and 2013.  The data fields changed over the period; we used data from the fourth quarter of 2013, the last available, for our analysis. The 2013 data have more benefit categories than the 2011 data and more information about limits on benefits than the 2012 data.  Our discussion with CMS staff suggested that the data became more complete as they were updated through the year, so we chose to use the fourth quarter.

The analysis is limited to plans for which the insurer reported that there were enrollees in the underlying product upon which the plan was based. Results are not enrollment weighted. Enrollment was reported only by product, so we do not know if there were actually enrollees in each of the plans associated with that product.  A plan was assumed not to offer coverage for a benefit category if the dataset left the coverage description blank for that category.  The analysis of benefit limits was conducted among plans that offered coverage for the benefit category analyzed, for example outpatient mental/behavioral health services.

1. In its application for a waiver, the state must show that the waiver would reduce average premiums for health insurance, increase enrollment in health insurance, stabilize the market for insurance coverage, stabilize premiums for applicants with pre-existing conditions, or increase choice of health plans.u00a0 AHCA, section 136. ↩︎
2. u00a0 Plans generally paid for complications from pregnancy, but not for the costs associated with a normal delivery. ↩︎
3. Roach, J., u201cDiscrimination and Mental Illness, Codified in Federal Law and Continued by Agency Practice,u201d 2016 Mich. St. Law Review 269, at 285:288. ↩︎

In 2017, Congress has been debating legislation, the American Health Care Act (AHCA), which would end the enhanced federal matching funds for the Affordable Care Act (ACA) Medicaid expansion and fundamentally alter the structure and financing of the Medicaid program. Specifically, it would cap and significantly reduce the amount of federal funding provided to states for Medicaid through a per capita cap or block grant. The Congressional Budget Office estimated that the AHCA as passed by the House would reduce federal Medicaid spending by $834 billion from 2017-2026 and reduce enrollment by 14 million by 2026 compared to projections in current law. The proposed Trump Administration budget for FY 2018 would have deeper Medicaid reductions.

The cap on federal funding would lock-in current state spending patterns that reflect historic Medicaid policy choices. Today, Medicaid is a state and federal partnership where the federal government sets core requirements for Medicaid and states administer the program; financing for Medicaid is shared by states and the federal government with no caps. Due to flexibility in the current law, states historically have made different Medicaid decisions related to coverage, scope of benefits, reimbursement rates and delivery system models. In response to a funding cap, each state would need to make budget decisions to fill in gaps in federal funding (through taxes or other budget cuts) or to restrict Medicaid programs.

All states could face challenges responding to federal Medicaid cuts and caps to varying degrees, but states with certain characteristics are more at risk. This analysis examines 30 factors in five groups that could be high risk factors affecting states’ ability to respond to federal Medicaid cuts and caps and identifies states ranked in the top five for each factor as high risk (ES-1). This analysis shows that more than 6 in 10 states rank in the top five for multiple risk factors. Eleven states rank in the top five for five or more risk factors (Alabama, Arizona, Florida, Georgia, Kentucky, Louisiana, Mississippi, New Mexico, South Carolina, Texas, and West Virginia).

States that adopted the Medicaid expansion have experienced gains in coverage and financing that are at risk under proposals like the AHCA that would end the enhanced federal match for the expansion. However, states that have not adopted the expansion would lose the option to access enhanced federal matching dollars for coverage in the future.

• Arkansas, Kentucky, Nevada and Oregon are among states that rank in the top five for multiple risk factors tied to the end of enhanced funding for expansion (i.e. large coverage gains, a high share of expansion enrollees, and a high share of expansion funding relative to the total). In total numbers, California, New York, Pennsylvania, Ohio and Illinois account for the largest number of enrollees in the expansion group (54%) and California, New York, Ohio, Michigan and Illinois account for the highest levels of federal expansion funding (52%).
• Non-expansion states would lose the future option to provide coverage to poor uninsured adults with enhanced federal dollars. Florida, Georgia and Texas have the largest number of uninsured residents who fall into the coverage gap (i.e. not eligible for Medicaid but have incomes below poverty so are not eligible for tax credits in the Marketplace) and the highest overall uninsured rates.

States with limited Medicaid programs and other challenging characteristics such as poor demographic indicators, poor health status, high cost health care markets and low state fiscal capacity could face more challenges in responding to per capita cap or block grant policies. States with limited Medicaid benefits or low provider reimbursement rates have less room to make further restrictions in benefits or to lower rates in response to reductions in federal financing. For these states, trade-offs within Medicaid would be difficult. States with an aging population, high levels of disability, a high share of people in health professional shortage areas or low per capita income may have higher demand for Medicaid services, but less capacity (especially with limited federal financing) to address those issues. In addition, because financing caps lock states into historic Medicaid decisions, states with limited programs and other risk factors will have a harder time adapting to future changes such as increased costs or changing demographics. One example of an emerging health issue is the opioid epidemic. Medicaid plays a central role in the nation’s effort to address the opioid epidemic through coverage of people struggling with opioid addiction and financing for states, limited funding could impede efforts to address this as well as other future health issues.

The findings below show states that rank in the top five for multiple risk factors within each of the five categories (beyond expansion):

• Medicaid Policy Choices: Alabama, Hawaii, Mississippi and Missouri have more than one restrictive Medicaid policies that would make it more challenging for them to implement additional program reductions (i.e. low eligibility levels, limited benefits, low provider reimbursement, high managed care penetration and/or a lower share of community based long-term care services).
• Demographics: Alaska, District of Columbia, Louisiana, New Mexico and Wyoming each have multiple demographic characteristics that indicate higher needs for Medicaid (i.e. high poverty, high unemployment, faster expected growth in the 85 year old population, high share of the population in rural areas, and / or high share of the population that is non-white).
• Health Status: Alabama, Arkansas, Louisiana, Mississippi and West Virginia have populations with multiple high health needs (i.e. overall poor health status, high share reporting a disability, high share reporting poor mental health, high opioid death rate and/or high rate of new HIV cases). Ohio, New Hampshire and West Virginia reported the highest opioid related drug overdoses at over 24 people in every 100,000.
• Revenue and Budget Choices: Alabama, Arizona, Idaho, Mississippi, New Mexico, South Carolina, Tennessee and West Virginia have multiple tax capacity challenges (i.e. low personal income, low total taxable resources, low tax effort or share of taxes relative to personal income, high Medicaid match rates, and/or low state and local spending per capita).
• Health Care Costs/Access: Alaska, Alabama, Florida, Louisiana, Massachusetts, Mississippi and New York have high cost health care markets and access challenges (i.e. high national health expenditures per capita, high employer sponsored insurance premiums, high share or people not seeking care due to cost, high share of the population in a shortage area and/or lower Medicaid physician participation).

While all states have risk factors to varying degrees, this analysis shows that more than 6 in 10 states rank in the top five for multiple risk factors. Eleven states rank in the top five for five or more risk factors (Alabama, Arizona, Florida, Georgia, Kentucky, Louisiana, Mississippi, New Mexico, South Carolina, Texas, and West Virginia). States that have multiple risk factors could face even more challenges making Medicaid program cuts or filling gaps in federal funding. Looking ahead, limiting the growth in federal Medicaid spending could force states to make difficult choices in their current programs and could also limit states ability to afford new drug therapies or other medical advances, adapt to changing demographics or make future investments to improve delivery systems or address broader health status issues in the future.

Tables with state-level data for each of the 30 risk factors considered in the analysis are included at the end of the report. Sources for data included in the tables and described in the report can be found in the appendix.

Introduction

Medicaid, the nation’s primary health insurance program for low-income and high-need Americans, is jointly financed by states and the federal government, but states administer Medicaid programs within broad federal rules. Under federal law, Medicaid provides an entitlement to coverage to individuals who are eligible for the program and a guarantee to federal matching dollars. Given the current financing structure of the program, Medicaid is both a budget item and a revenue item in state budgets. State Medicaid policy choices (including the ACA coverage expansion, other eligibility levels, scope of benefits, reimbursement rates and delivery system models) as well as other factors such as demographics, health needs, health care markets, and state fiscal capacity affect Medicaid spending. Due to all of these factors, Medicaid coverage and financing vary significantly across states (Figures 1 and 2).

Figure 1. State Characteristics

This brief, builds on earlier work that examined the factors contributing to variation in Medicaid spending from 2012. This update reflects changes including the implementation of the ACA and a slowly improving economy. In 2017, Congress has been debating the AHCA, legislation that could eliminate enhanced federal matching funds for the ACA Medicaid expansion and fundamentally change the structure and financing of the overall program by capping and significantly reducing the amount of federal funds provided to states. This brief examines 30 key factors that contribute to variation in Medicaid per enrollee spending and highlights states that may be at particularly high risk under federal cuts or caps in Medicaid funding. We identify “high risk” states as ranking in the top five for each factor. While we use this ranking construct, there may not be significant differences between states that follow in rank order. The brief has findings in five key areas:

• Medicaid policy choices (including implementation of the ACA Medicaid expansion);
• Demographics;
• Health status;
• Available tax revenues and state budget choices; and
• Health care markets (costs and access).

Table 1 includes state level data to show expansion status, Medicaid per enrollee spending for FFY 2014 and number of state residents. The appendix provides sources for data included in the tables and described in the report.

Key Findings

1.  Medicaid Policy Choices

Medicaid ACA Expansion

The ACA expanded Medicaid coverage to nearly all adults up to 138% of the federal poverty level (FPL) and provided states with enhanced federal financing for this new coverage; however, the Supreme Court ruling that effectively gave states the option to adopt the expansion resulted in even greater variation in Medicaid programs across states. Following the implementation of the ACA in 2014, millions of people have enrolled in new coverage options (Medicaid and Marketplace), and the uninsured rate for the non-elderly had dropped to a historic low of 10% by early 2016. The proposed elimination of federal enhanced matching funds for the ACA Medicaid expansion included in the AHCA would have different implications for states based on the decision to implement the expansion and the effects of that decision in a given state. States that expanded have experienced benefits tied to coverage gains and financing that could be at risk, while non-expansion states risk the future loss of Medicaid enhanced matching funds for expansion (Figure 3 – states in multiple categories are shown in orange).

Expansion States. Policy proposals that would end the enhanced federal matching funds for the ACA Medicaid expansion could have significant implications for coverage and financing in expansion states. To date, 32 states (including DC) adopted the Medicaid expansion. Coverage gains were particularly large among low-income people living in states that expanded Medicaid. At the start of 2016, there were 14.4 million adults in the Medicaid expansion group, including 11 million who were “newly” eligible due to the ACA expansion. Overall, expansion enrollment accounted for about 20% of all Medicaid enrollment. From 2014-2016, the federal government financed 100% of the costs of those newly eligible under the expansion. Under the law, this rate phased down to 95% in 2017 and gradually phases down to 90% by 2020. Federal funding for the expansion group totaled $68.2 billion in FY 2015 (about 21% of all federal Medicaid funding). In total numbers, California, New York, Pennsylvania, Ohio and Illinois account for the largest number of enrollees in the expansion group (54%) and California, New York, Ohio, Michigan and Illinois account for the highest levels of federal expansion funding (52%).

• In Oregon, New York, Arkansas, West Virginia, and Kentucky, Medicaid expansion enrollees accounted for 35% or more of total enrollment (Figure 4).
• In Washington, Oregon, Nevada and Kentucky, expansion funding accounts for at least 40% of all federal Medicaid funds.
• Nevada, Kentucky, California, New Hampshire and Arkansas experienced the largest percentage point reduction in the uninsured from 2013 to 2015.

Figure 4. Affordable Care Act (ACA) Expansion

Eight of the expansion states (Arkansas, Arizona, Illinois, Indiana, Michigan, New Hampshire, New Mexico and Washington) have legislation requiring them to reduce or eliminate the expansion if the federal match rate is reduced. Other states are likely to eliminate or scale back their expansion coverage due to the increased cost if federal funding is reduced. As a result, many expansion adults in these states would likely become uninsured since they would not be able to afford other coverage options. Adults covered by Medicaid typically work in low-wage jobs that do not provide health coverage or are family caregivers, and many have chronic conditions or disabilities. Increases in uninsured rates among these adults would increase strains on other parts of state health care systems, including community health centers and hospitals.

Non-Expansion States. The 19 states that have not expanded would lose the option to expand in the future and access to enhanced federal funds to support their capacity to cover low-income uninsured adults. Most of the non-expansion states (12 of 19) limit Medicaid eligibility for parents to less than half the poverty level and other low-income adults are not eligible regardless of income, except in Wisconsin. Many poor parents and other adults in these states fall into a coverage gap since they do not qualify for Medicaid, but have incomes below 100% of poverty so cannot access tax credits to purchase Marketplace coverage. The ongoing coverage gap for low-income adults in these states limits state capacity to achieve overall improvements in population health. Texas, Florida, Georgia, North Carolina, and South Carolina are the states with the highest number of adults who fall into the coverage gap.

Table 2 includes state level data related to the ACA Medicaid expansion and source information can be found in the appendix.

Other Medicaid Policy Choices

States administer Medicaid within broad federal rules. The federal government sets core requirements, but states have flexibility to determine eligibility levels, benefits, provider payments, and delivery systems. There is a great deal of variation across states in how their Medicaid programs are structured, and therefore differences in spending on the program due to policy choices made by states. Under a per capita cap or block grant policy, states could be locked into historic decisions about their Medicaid programs.

States with more limited Medicaid programs in terms of eligibility, benefits and payment rates could be at higher risk with reductions or caps in federal financing because it could be more difficult for these states to find savings if some core federal requirements remain in place. Moreover, nearly all community based long-term care services are an optional service in Medicaid, so states that have lower levels of home and community-based services (HCBS) may find it difficult to expand these services to shift utilization and achieve savings from reductions in more costly institutional care in the face of federal Medicaid cuts. The AHCA repeals the 6% enhanced federal funding for the Community First Choice Option to provide attendant care services as of 2020, which would result in a $12 billion decrease in federal funding over 10 years and further hinder efforts to increase community integration. In addition, states that have already moved enrollees to managed care would not be able to recoup some of the one-time savings that some states experience in that transition.

Figure 5 identifies states with restrictive Medicaid programs that would face particular challenges to achieving increased savings within their programs in response to reductions in federal funding. States in multiple categories are shown in orange.

Eligibility. Federal law requires states to cover certain population groups up to minimum income eligibility levels in order to receive federal matching funds. States can expand coverage beyond federal minimum levels and receive federal matching funds. As of January 2017, the median eligibility levels were 255% of the federal poverty level (FPL) for children and 138% FPL for parents and adults without dependent children. Eligibility levels for adults in non-expansion states are much lower for parents and Wisconsin is the only non-expansion state to provide Medicaid coverage to childless adults. Alabama, Texas, Missouri, Idaho and Mississippi have the lowest eligibility levels for parents. The median eligibility thresholds for parents in non-expansion states is 44% FPL (Figure 6).

Figure 6. Medicaid Policy

Benefits. States participating in Medicaid must cover a core set of benefits, but states can also receive federal matching funds for coverage of “optional benefits.” All states cover some optional benefits, but the scope of state benefit packages varies widely across states. States also have flexibility in determining the amount, duration and scope of the benefits they offer. Adult dental benefits can be used as a proxy to understand the variation across state benefit packages. While almost all states (46) and DC currently provide some dental benefits for adults in Medicaid, the scope of Medicaid adult dental benefits varies widely by state. As of February 2016, 15 states provided extensive adult dental benefits, 19 states provided limited dental benefits, and 13 states covered only dental care for pain relief or emergency care for injuries, trauma, or extractions. Four states provided no dental benefits at all. Like other optional Medicaid services, adult dental benefits are often cut when states face budget pressures.

Payment Rates/Physician Fees. States largely determine provider payments within limited federal requirements. Federal law requires that payments be consistent with efficiency, economy, quality and access, and safeguard against unnecessary utilization. For physicians and other providers, states are required to pay rates that are sufficient to ensure access equal to the rest of the area population. In 2014, Medicaid paid physicians on average 66% of Medicare rates and 59% of what Medicare pays for primary care. Relative to what other states reimburse for physicians, Rhode Island, New Jersey, Michigan, California and Missouri have the lowest physician fee index.

Managed Care. Today, 39 states contract with managed care organizations (MCOs) and risk-based managed care is the dominant delivery system in Medicaid. Hawaii, Tennessee, Iowa, New Hampshire, Kansas and New Jersey all report that at least 95% of enrollees are in MCOs. Sometimes, when states first transition to managed care arrangements, they can see some savings relative to fee-for-service arrangements, but savings are not sustained not over time. States that have already transitioned to managed care have less ability to use this as a mechanism to reduce future per enrollee costs.

Long-Term Care Setting. Medicaid covers a continuum of long-term care service settings. Driven by requirements under the Americans with Disabilities Act (ADA), consumer preferences and efforts to control long-term care costs, states continue to work on reorienting their Medicaid long-term care delivery systems away from institutional settings and toward more community-based services. Overall, more than half of spending (55%) for long-term care is for HCBS, but this varies by state from a high of 82% in Oregon to 31% in Mississippi.

Table 3 includes state level data related to Medicaid policies and source information can be found in the appendix. 

2.  Demographics

Demographic characteristics vary across states. These characteristics affect demand for Medicaid as well as public health services more broadly. States with high rates of poverty, unemployment, and uninsured could have higher demands for Medicaid. Similarly, because Medicaid is the largest provider of long-term services and supports, states with higher shares of elderly as well as states facing the fastest growth projections in their population over 85 years old could face increased demand for Medicaid. In addition, Medicaid plays an important role in providing coverage to people of color as well as people in rural areas. States with demographic factors that contribute to high demand for Medicaid could be disproportionately impacted by cuts or caps in federal Medicaid funding compared to other states. High demand and limited funding could make it challenging for these states to meet the needs of the residents in their state. Individuals in poverty or who are unemployed are unlikely to have another source of health coverage and those who are uninsured are at risk for not getting needed health services.

Figure 7 highlights states with demographic characteristics that lead to increased need for Medicaid. States in multiple categories are shown in orange.

Poverty. Many government assistance programs are targeted to help low-income families. Programs administered or supported by the United States Department of Health and Human Services (HHS) use the department’s federal poverty guidelines. The annual federal poverty level in 2017 is $12,060 for an individual and $20,420 for a family of three. In 2015, 14% of people living in the U.S. had incomes below the poverty level. The percent of the population living below poverty varies by state, ranging from less than 10% of residents in Alaska, Connecticut, Minnesota, New Hampshire and Utah up to 20% in Kentucky and New Mexico.

Unemployment. During an economic downturn, individuals lose jobs, incomes drop, and state revenues decline at the same time that demand for public programs such as Medicaid, cash assistance, and food stamps increase. During the most recent recession, unemployment peaked at 10% in October 2009. The economy has been improving since then and unemployment rates have been stable at less than 5% since May 2016, but rates vary across states. As of April 2017, the national unemployment rate was 4.4%. Colorado, Hawaii, Maine, Nebraska, New Hampshire, North Dakota, and South Dakota had unemployment rates at 3% or less, while Alaska and New Mexico had rates greater than 6%.

Age. Demand for health care services varies by age. For example, those over the age of 65 are more likely to need long-term care services than children, who generally require fewer medical services and mostly preventive care and acute care. In 2015, the elderly accounted for 15% of the population and the population 85 years of age and older accounted for 2%.¹  From 2015 to 2030, the percent of the population 85 years and older is expected to grow by 41% with 21 states projecting growth in the 85 and older population of more than 50% (Figure 8).

Figure 8. Demographics

Rural. States with large rural populations tend to have poorer residents with complex and costly health needs, which are often exacerbated by workforce shortages and large geographic distances between patients and providers. Medicaid plays a central role in helping to fill gaps in private coverage in rural areas. States including Montana, North Dakota, and Wyoming have all of their populations living in rural areas.

Race. Communities of color are more likely to face disparities in health and access to health care and Medicaid plays an important role in improving coverage for persons of color. In seven states (California, District of Columbia, Hawaii, Maryland, Nevada, New Mexico, and Texas), the share of the population that is non-white is 50% or more.

Table 4 includes state level data related to demographic characteristics and source information can be found in the appendix.

3.  Health Status

Health status characteristics vary across states and these characteristics affect demand for Medicaid as well as public health services more broadly. Medicaid plays an important role in serving individuals with disabilities, addressing public health crises, and providing mental health care. Medicaid also plays a central role in helping states address the opioid crisis and the HIV epidemic. Cuts or caps in federal Medicaid financing could increase the challenges in addressing the opioid and HIV epidemics and providing care for individuals with disabilities. Figure 9 shows states that have particularly high health needs across certain measures; states in multiple categories are shown in orange.

Overall Health Ranking. America’s Health Rankings^® Annual Report provides annual state-by-state rankings based on behaviors, community and environment, policy, clinical care, and outcomes data. Based on these data, Hawaii, Massachusetts, Connecticut, Minnesota and Vermont are the healthiest states and Mississippi, Louisiana, Arkansas, Alabama and Oklahoma have the greatest challenges.

Disability. Medicaid covers more than 3 in 10 nonelderly adults with disabilities, providing a broad range of medical and long-term care services that enable people with disabilities to live and work in the community. There are a number of factors that determine the need for health and long-term care services in states, including the prevalence of disability and chronic conditions as well as other indicators of health. In 2015, 13% of the non-institutionalized population reported having a disability. Sixteen states had 15% or more of noninstitutionalized people reporting a disability (Figure 10). In FY 2014, Medicaid enrollees with a disability accounted for 40% of Medicaid spending, but only 14% of Medicaid enrollment; the spending per enrollee for these individuals was three times as much as the average spending per all enrollees.² 

Figure 10. Health Status

Mental Health. The Medicaid program covers a disproportionate share of individuals with behavioral health conditions. More than one in four adults with Medicaid (27%) have a mental illness. The Medicaid program serves as a safety net for many low-income individuals with behavioral health conditions by facilitating access to and financing numerous services, including clinical services, case management, prescription medication and rehabilitative services. Over one-third of adults nationally (35%) report poor mental health.

Opioid Deaths. The opioid epidemic is increasing among Americans with addiction to heroin and prescription painkillers. Medicaid plays a central role in the nation’s effort to address the opioid epidemic through coverage of people struggling with opioid addiction and financing for states. In 2015, more than 10 people in every 100,000 died from an opioid related drug overdose nationally and opioid overdoses accounted for 63% of all drug overdoses. West Virginia, New Hampshire and Ohio reported the highest opioid related drug overdoses at over 24 people in every 100,000.

HIV. Medicaid is the largest source of coverage for individuals with HIV. By covering more people with HIV and increasing state capacity to provide treatment to individuals with HIV, Medicaid plays a key role in curbing the epidemic. In 2015, the new HIV diagnoses rate among adults and adolescents was 14.7 people for every 100,000.

Table 5 includes state level data related to health status and source information can be found in the appendix.

4.  Available Revenues And State Budget Choices

States vary in the amount of revenue resources available and in how they tap into those resources to pay for public services and programs. States also make different decisions about how revenue is allocated across different areas of spending. Medicaid is an important revenue source for states, but also a major source of spending. Figure 11 highlights states with lower tax bases, lower tax collections and higher federal Medicaid match rates, which would be at higher risk with reductions or caps in federal financing because it would be more difficult for these states to offset the loss of federal funds with state funds. States in multiple categories are shown in orange.

Tax Base. A states’ tax base provides a measure of resources that may be available to a state. Two measures of tax base include personal income and total taxable resources. Per capita personal income³  is currently used to determine the Federal Medical Assistance Percentage (FMAP) for Medicaid. States with low personal income per capita have a higher FMAP. Personal income per capita does not include all potential taxable income sources in a state.⁴  In 2014, personal income per capita was $46,049 nationally (Figure 12). Total taxable resources (TTR) is an alternative and more comprehensive measure of tax base designed to better account for a state’s available resources to provide public services.^5 ,6  Mississippi, West Virginia, South Carolina, Idaho and New Mexico had the lowest personal income per capita in 2014. Mississippi, Arizona, Idaho, South Carolina and Alabama had the lowest TTR in 2014.

Figure 12. Available Revenues and State Budgets

Tax Collections. Tax collections are one measure of how a state taps into its available resources. In 2014, tax collections per capita averaged $4,675 across the U.S.⁷  This brief examines tax collections as a percent of personal income as a proxy for a more representative measure of tax effort, that is, how much a state collects in taxes relative to its tax capacity. Tax collections as a share of personal income averaged 10%. Tennessee, Florida, Alabama, South Dakota and Oklahoma had the lowest tax collections as a share of personal income.

State and Local Spending Per Capita. Because states must balance their budgets annually, state spending is a function of how much revenue is collected. Average state and local spending per capita in 2014 was $8,489. Idaho, Arizona, Georgia, Nevada and Tennessee had the lowest state and local spending per capita from all sources.

Federal Medical Assistance Percentage (FMAP). As noted above, the FMAP is based on a state’s relative personal income per capita. States with lower relative personal income have a higher FMAP. In FY 2018, the FMAP ranged from a floor of 50% to a high of 76% in Mississippi.

Table 6 includes state level data related to income, tax, spending data and FMAP variables and source information can be found in the appendix.

5.  Health Care Markets (Costs and Access)

Medicaid programs purchase services from the private market as other insurers do. Therefore, some of the variation in Medicaid spending is due to differences in health care markets and the ability to access care as well as the number of providers and health care facilities in a given state. The share of the population living in health professional shortage areas and the share of providers accepting Medicaid also varies by state and can contribute to overall variation in per enrollee costs. Figure 13 highlights states with high health care costs and limited access, which could be at higher risk under reductions or caps in federal financing because it would become increasingly difficult to purchase services in a high cost market or make improvements in access to care. It also highlights states with low Medicaid physician participation, which would be at risk under Medicaid cuts or caps because it would be difficult to increase provider rates beyond the caps to promote additional provider participation. States in multiple categories are shown in orange.

Health Care Costs. Medicaid spending is, in part, affected by the cost of procuring health care in the state market. Differences in the cost of obtaining health care in a particular state affects the amount state Medicaid programs have to spend in order to purchase services. Health care costs typically outpace inflation. In 2009, national health expenditures (NHE) per capita were $6,815. Half of all U.S. residents and 56% of non-elderly residents are covered by employer-sponsored insurance (ESI) plans. In 2015, the average premium cost for a family (including the employee and employer shares) for employer-based coverage was $17,322 (with 73% paid by the employer).

Access to Care. Access varies across states due to factors such as provider availability, geography and population density, payment, and other local factors. The presence of access barriers may influence Medicaid beneficiaries’ ability to access services and, in turn, state Medicaid spending. Nationally, 13% of the population reported that they had not seen a doctor because of cost. This was most likely to occur in Mississippi, Texas, Florida, Alabama, and South Carolina. Areas with a documented shortage of providers are designated as Health Professional Shortage Areas (HPSAs) by the Health Resources and Services Administration (HRSA). As of December 2016, HRSA had designated 6,626 primary care HPSAs across the country, affecting over 66 million people. Nationally, 20% of the population lives in a primary care HPSA. Mississippi, New Mexico, Louisiana, Arizona and Alabama have the highest share of their populations in primary care HPSAs (Figure 14).

Figure 14. Health Care Markets (Cost and Access)

Physician Participation. Nationally, 69% of all office-based physicians accept new Medicaid patients. The percentage of physicians accepting new Medicaid patients varies by state, ranging from 39% in New Jersey to 97% in Nebraska. In one-quarter of states (14), more than 85% of physicians accept new Medicaid patients.

Table 7 includes state level data related to health care market costs and access, and source information can be found in the appendix.

Looking Ahead

Today, Medicaid is a state and federal partnership. The program is largely administered by states within broad federal rules. States have flexibility to make policy choices related to eligibility, benefits and how to deliver and pay for care. States and the federal government share financing for Medicaid and states have a guarantee to federal matching dollars with no set limit. Under current law, Medicaid spending per enrollee varies across states due to a complex array of factors that involve state policy choices, residents’ needs for public health services as well as health and long-term care, states’ abilities to raise revenue and collect revenue, Medicaid policy choices, and the underlying health care market in a state.

Congress continues to debate and consider legislation to eliminate the enhanced match for the Medicaid ACA expansion and cut and cap federal Medicaid financing through a block grant or per capita cap. The cap on federal funding would lock-in current state spending patterns that reflect historic Medicaid policy choices. All states could face challenges responding to federal Medicaid cuts and caps, but states with certain characteristics are more at risk.

Challenging demographics, poor health status, low tax capacity and state spending, high health care costs and poor access are key factors that would place states at higher risk from a per capita cap or block grant. If states are locked in to prior policy choices, states with limited eligibility and benefits, lower provider rates and those that have already moved to capitated care arrangements could be at higher risk with federal Medicaid spending caps because these states would have few options to reduce the scope of the program or generate new efficiencies. States that have seen the largest gains in coverage and those that have high shares of federal Medicaid dollars from the expansion could see the biggest losses if the ACA Medicaid expansion funds are eliminated, while other states that have not expanded, could lose the future opportunity to expand and see similar gains in coverage and financing.

This analysis of 30 factors in 5 groups shows that all states have risk factors to varying degrees; more than 6 in 10 states rank in the top five for multiple risk factors. Eleven states rank in the top five for five or more risk factors (Alabama, Arizona, Florida, Georgia, Kentucky, Louisiana, Mississippi, New Mexico, South Carolina, Texas, and West Virginia). Looking ahead, limiting the growth in federal Medicaid spending could force states to make difficult choices in their current programs and could also limit states’ ability to afford new drug therapies or other medical advances, adapt to changing demographics or make future investments to improve delivery systems or address broader health status issues.

Table 1. State Characteristics

Table 2. Affordable Care Act (ACA) Expansion

 

Table 3. Medicaid Policy

 

Table 4. Demographics

 

Table 5. Health Status

 

Table 6. Available Revenues and State Budgets

 

Table 7. Health Care Markets (Cost and Access)

Appendix: Table Notes and Sources

Table 1: State Characteristics

Expansion Status, as of January 1, 2017Source: Kaiser Family Foundation’s State health Facts, Status of State Action on the Medicaid Expansion Decision, as of January 1, 2017.

Share of Medicaid Spending, by Enrollment Group, FFY 2014Source: Kaiser Family Foundation estimates based on analysis of data from the FFY 2014 Medicaid Statistical Information System (MSIS) and CMS-64 reports. Because FY 2014 data was missing some or all quarters for some states, we adjusted the data using secondary data to represent a full fiscal year of enrollment.

Total Number of Residents, 2015Note: Population numbers are rounded to the nearest 100.Source: Kaiser Family Foundation estimates based on the Census Bureau’s March 2016 Current Population Survey (CPS: Annual Social and Economic Supplement).

Table 2: ACA Expansion

Expansion Enrollees, Number and as a Share of Total Medicaid Enrollees, January-March 2016Note: Louisiana expanded Medicaid on July 1, 2016 and thus has no expansion enrollment in January-March 2016. Data is not available for North Dakota.Source: Kaiser Family Foundation analysis of Medicaid spending and enrollment data collected from the Centers for Medicare and Medicaid Services (CMS) Medicaid Budget and Expenditure System (MBES).

Federal Expansion Funding, Amount and as a Share of all Federal Medicaid Spending, FFY 2015Note: Alaska expanded on September 1, 2015 and thus reported expansion data for one month in FFY 2015. Louisiana (7/1/2016) and Montana (1/1/2016) expanded after FFY 2015.Source: CMS, Medicaid Budget and Expenditure System (MBES) Expenditure Reports, December 2016.

Percentage Point Change in Uninsured Rate for Nonelderly, 2013-2015Source: Kaiser Family Foundation estimates based on the Census Bureau’s March 2014 and March 2016 Current Population Survey (CPS: Annual Social and Economic Supplements).

Nonelderly Uninsured Rate, 2015Source: Kaiser Family Foundation estimates based on the Census Bureau’s March 2014, March 2015, and March 2016 Current Population Survey (CPS: Annual Social and Economic Supplements).

Number of Adults in the Medicaid Coverage Gap, 2016Note: Wisconsin covers adults up to 100% FPL in Medicaid under a waiver but did not adopt the ACA expansion. No data is available for Maine.Source: Kaiser Family Foundation analysis based on 2016 Medicaid eligibility levels and 2016 Current Population Survey.

Table 3: Medicaid Policy

Eligibility Income Levels for Parents, as a Percent of the Federal Poverty Level, as of January 1, 2017Notes: Eligibility levels are based on the FPL for a family of three, which is $20,420.Sources: Medicaid and CHIP Eligibility, Enrollment, Renewal, and Cost Sharing Policies as of January 2017: Findings from a 50-State Survey, Kaiser Family Foundation, January 2017. Based on a national survey conducted by the Kaiser Commission on Medicaid and the Uninsured with the Georgetown University Center for Children and Families, 2017.

Scope of Medicaid Adult Dental Benefits, as of February 2016Source: Center for Health Care Strategies, Inc. (CHCS), Medicaid Adult Dental Benefits: An Overview, (Hamilton, NJ: CHCS, May 2017), https://www.chcs.org/resource/medicaid-adult-dental-benefits-overview/.

Medicaid Physician Fee Index, 2014Note: Tennessee does not have a Medicaid fee-for-service program and so data is not applicable.Source: Stephen Zuckerman, Laura Skopec, and Kristen McCormack, “Reversing the Medicaid Fee Bump: How Much Could Medicaid Physician Fees for Primary Care Fall in 2015?,” Urban Institute, December 2014.

Shared of Medicaid Population in a Managed Care Organization (MCO), as of July 1, 2016Source: Vernon K. Smith, Kathleen Gifford, Eileen Ellis, and Barbara Edwards, Health Management Associates; and Robin Rudowitz, Elizabeth Hinton, Larisa Antonisse and Allison Valentine, Kaiser Commission on Medicaid and the Uninsured. Implementing Coverage and Payment Initiatives: Results from a 50-State Medicaid Budget Survey for State Fiscal Years 2016 and 2017, Kaiser Family Foundation, October 2016.

Percent of Long-Term Care Spending for Home and Community Based Services (HCBS), FY 2015Note: Data do not include expenditures for managed care programs in California and North Carolina. Percent HCBS is not calculated for these states because a significant portion of data are missing.Source: Truven, Medicaid Expenditures for Long-Term Services and Supports (LTSS) in FY 2015, April 14, 2017

Table 4: Demographics

Percent of the population below 100% FPL, 2015Note: The U.S. Census Bureau’s poverty threshold for a family with two adults and one child was $19,078 in 2015.Source: Kaiser Family Foundation estimates based on the Census Bureau’s March 2016 Current Population Survey (CPS: Annual Social and Economic Supplements).

Unemployment Rate, as of April 2017Source: Bureau of Labor Statistics (BLS), Regional and State Employment and Unemployment (Monthly), Table 1, Civilian labor force and unemployment by state and selected area, seasonally adjusted, April 2017; and BLS Employment Situation News Release, State Employment and Unemployment Summary – April 2017, May 19, 2017.

Projected Change in persons 85+, from 2015 to 2030Source: U.S. Census Bureau, Population Division, Interim State Population Projections, 2005, compiled by the US Administration on Aging.

Percent of Nonelderly Population in Rural Areas, 2015Source: Kaiser Family Foundation analysis based on the 2015 American Community Survey 1-Year Estimates.

Percent of Total Population that is Non-White, 2015Note: Non-White includes Black, Hispanic, Asians, NHOPIs, American Indians and persons of two or more races.Sources: Kaiser Family Foundation estimates based on the Census Bureau’s March 2016 Current Population Survey (CPS: Annual Social and Economic Supplements).

Table 5: Health Status

Overall State Health Ranking, 2016Note: The Overall Health Status Score is compiled by the United Health Foundation and is a weighted standard deviation relative to the US value. It is based on a series of measures related to health behavior, community and environment, policy, clinical care, and health outcomes. For a complete list of measures and methodology, see http://assets.americashealthrankings.org/app/uploads/ahr16-complete-v2.pdf. The District of Columbia is not ranked.Source: United Health Foundation, America’s Health Rankings: 2016 Annual Report; May 25, 2016.

Percentage of Non-Institutionalized Population Who Reported a Disability, 2015Source: Erickson, W., Lee, C., von Schrader, S. (2017). Disability Statistics from the 2014 American Community Survey (ACS). Ithaca, NY: Cornell University Employment and Disability Institute (EDI). Retrieved June 7, 2017 from www.disabilitystatistics.org.

Percent of Adults Reporting Poor Mental Health, 2015Source: Kaiser Family Foundation analysis of the Centers for Disease Control and Prevention (CDC)’s Behavioral Risk Factor Surveillance System (BRFSS) 2013-2015 Survey Results.

Age-Adjusted Opioid Overdose Deaths Rates, per 100,000Source: Kaiser Family Foundation analysis of Centers for Disease Control and Prevention (CDC), National Center for Health Statistics. Multiple Cause of Death 1999-2015 on CDC WONDER Online Database, released 2016. Data are from the Multiple Cause of Death Files, 1999-2015, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative Program. Accessed at http://wonder.cdc.gov/mcd-icd10.html on March 2, 2017.

New HIV Diagnoses, Among Adults and Adolescents, per 100,000 Population, 2015Source: Centers for Disease Control and Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention (NCHHSTP) AtlasPlus accessed February 2017.

Table 6: Taxes

Personal Income Per Capita, 2014Source: Bureau of Economic Analysis, state personal income accounts, 2014.

Total Taxable Resources Per Capita, 2014Source: U.S. Department of Treasury, Total Taxable Resources, accessed April 2017.

Tax Collections as a Percent of Personal IncomeSource: KFF analysis of Bureau of Economic Analysis, state personal income accounts, 2015; and U.S. Census Bureau, 2014 State & Local Government Finance, 2014.

Total State and Local Spending per Capita, 2014Source: U.S. Census Bureau, 2014 State & Local Government Finance, 2014.

Federal Medical Assistance Percentage (FMAP) for Medicaid, 2018Source: Federal Register, November 15, 2016 (Vol 81, No. 220), pp 80078-80080.

Table 7: Costs and Access

Health Care Expenditures per Capita, 2009Source: Centers for Medicare & Medicaid Services, Office of the Actuary, National Health Statistics Group. National Health Expenditure Data: Health Expenditures by State of Residence, December 2011. U.S. Population by State, 1991-2009 obtained from the U.S. Bureau of the Census, February, 2011.

Average Family Premium per Enrolled Employee for Employer-Based Health Insurance, 2015Source: Agency for Healthcare Research and Quality, Center for Financing, Access and Cost Trends. Medical Expenditure Panel Survey (MEPS) -Insurance Component, 2013-2015, Tables II.D.1, II.D.2, II.D.3 available at: Medical Expenditure Panel Survey (MEPS).

Share of Adults Reporting not Seeing a Doctor Due to Cost, 2015Source: Kaiser Family Foundation analysis of the Center for Disease Control and Prevention (CDC)’s Behavioral Risk Factor Surveillance System (BRFSS) 2013-2015 Survey Results.

Share of Population in Health Profession Shortage Area for Primary Care, as of December 2016Sources: Bureau of Health Workforce, Health Resources and Services Administration (HRSA), U.S. Department of Health & Human Services, Designated Health Professional Shortage Areas Statistics: Designated HPSA Quarterly Summary, as of December 31, 2016 and U.S. Census Bureau, State Population Totals Datasets: 2010-2016.

Physicians Accepting new Medicaid Patients, 2013Source: Hing et al., “Acceptance of new Patients with Public and Private Insurance b Office-based Physicians: United States, 2013,” NCHS Data Brief, No. 195, March 2015, CDC, USDHHS.

1. KFF analysis of the March 2016 Current Population Survey, Annual Social and Economic Supplement. ↩︎
2. KFF estimates based on analysis of data from the FFY 2014 Medicaid Statistical Information System (MSIS) and Urban Institute estimates from CMS-64 reports. ↩︎
3. Personal income is the income that is received by persons from all sources. It is calculated as the sum of wage and salary disbursements, supplements to wages and salaries, proprietors’ income with inventory valuation and capital consumption adjustments, rental income of persons with capital consumption adjustment, personal dividend income, personal interest income, and personal current transfer receipts, less contributions for government social insurance. This measure of income is calculated as the personal income of the residents of a given area divided by the resident population of the area. BEA uses the Census Bureau’s annual midyear population estimates. http://www.bea.gov/regional/definitions/nextpage.cfm?key=Per%20capita%20personal%20income%20%28dollars%29. ↩︎
4. Items excluded from person income include profits retained for investment purposes by corporations or other business entities and business or commuter income earned in the state by out of state residents, which can be influential in areas with large commuter populations, i.e. New York and New Jersey. ↩︎
5. TTR estimates are currently used to allocate funds for the Community Mental Health Services and Substance Abuse Prevention and Treatment block grants. ↩︎
6. Total Taxable Resources (TTR) for the District of Columbia was calculated to be over $101,000. However, because the District of Columbia does not have the same legal right as the states to tax certain resources, using the same methodology to derive TTR estimates for the District of Columbia is flawed. http://www.treasury.gov/resource-center/economic-policy/Documents/wpnewm.pdf. ↩︎
7. U.S. Census Bureau, 2014 State & Local Government Finance, 2014. ↩︎