Note: This analysis was updated on November 25th to reflect the October version of the 2025 CMS Landscape file.Over the last decade, Medicare Advantage, the private plan alternative to traditional Medicare, has taken on a prominent role in the Medicare program. In 2024, nearly 33 million Medicare beneficiaries are enrolled in a Medicare Advantage plan, more than half, or 54%, of the eligible Medicare population.

Despite concerns that modifications to the payment formula and higher utilization would impact the number of Medicare Advantage plans offered in 2025, the Medicare Advantage market appears to be relatively stable. While Medicare Advantage insurers have made some adjustments in their offerings, the average Medicare beneficiary has a choice of more than 30 Medicare Advantage prescription drug (MA-PD) plans, and virtually all plans provide multiple extra benefits like vision, hearing and dental benefits, similar to last year.

This brief provides an overview of the Medicare Advantage plans that are available for 2025 and key trends over time. This analysis uses data from the CMS Landscape files. In general, this brief refers to individual Medicare plans available for general enrollment, excludes Special Needs Plans (SNPs), except where noted, and excludes employer plans. (See methods for more details.) A second, companion analysis, describes premiums and benefits that are available for 2025 Medicare Advantage plans and over time.

Medicare Advantage Highlights for 2025

• The average Medicare beneficiary will have the option of 34 Medicare Advantage prescription drug (MA-PD) plans in 2025, 2 fewer than the 36 options available in 2024. Across all plans for individual enrollment, including those with and without prescription drug coverage, the average beneficiary has 42 options in 2025, compared to 43 options in both 2023 and 2024.
• The number of plans available to the average beneficiary varies across states. In 27 states and Puerto Rico, the average beneficiary has a choice of fewer plans in 2025 than in 2024, while in 16 states and DC, the average beneficiary has a choice of more plans, and in six states the number of plans available, on average, stayed the same.
• Nearly one-third of Medicare beneficiaries (32%) live in a county with more than 50 Medicare Advantage plans available in 2025, up from 7 percent in 2020, and similar to 2024 (33%). Less than 0.5 percent live in a county with no plans available.
• The average Medicare beneficiary can choose among plans offered by 8 firms in 2025, the same as in 2024. Three new insurers entered the Medicare Advantage market in 2025, while eight firms exited the market in 2025.
• Major insurers are both expanding into new counties and exiting others. For example, Humana is entering 12 new counties and exiting 70 counties, while UnitedHealthcare is entering 42 new counties and exiting 38 counties. Both insurers are offering plans in nearly 90% of all U.S. counties.
• About 5% of current Medicare Advantage enrollees in individual MA-PDs (or about 1.4 million people) are in a plan that has been terminated for 2025, while up to 7% of enrollees in MA-PDs (or as many as 1.8 million people) may be affected by a consolidation, meaning they are in a plan where some portion of the 2024 enrollment will be automatically assigned to a different plan in 2025.
• More than half (58%) of Medicare beneficiaries live in a county (1,188) where at least one firm is offering 10 or more plans for individual enrollment.

Plan Offerings in 2025

Number of Plans

Number of Plans Available to Beneficiaries. For 2025, the average Medicare beneficiary will have access to 34 Medicare Advantage prescription drug (MA-PD) plans, just 2 fewer than the 36 in 2024 (Figure 1). Across all plans for individual enrollment, including those with and without prescription drug coverage, the average beneficiary has 42 options in 2025, compared to 43 options in both 2023 and 2024. Since 2018, the number of plans available to the average beneficiary has doubled. These numbers exclude employer- or union-sponsored group plans, Special Needs Plans (SNPs), PACE plans, cost plans, and Medicare-Medicaid plans (MMPs) that are only available to select populations.

Total Number of Plans. In total, 3,719 Medicare Advantage plans are available nationwide for individual enrollment in 2025 – a 6% decrease from the number of plans (240 fewer plans) offered in 2024 (Appendix Figure 1).

HMOs account for more than half (56%) of all Medicare Advantage plans offered in 2025 but have declined as a share of all Medicare Advantage plans since 2017 (71% of plans), while during this period local PPOs rose as a share of all plans (Appendix Figure 1). During this period, the share of plans that are local PPOs increased from 24% to 43%. The share of plans that are regional PPOs has slowly declined from around 3% of plans offered in 2017 to 1% in 2025.

While many employers and unions also offer Medicare Advantage plans to their retirees, no information about these 2025 plan offerings is made available by CMS to the public during the Medicare open enrollment period. Employer and union plans are administered separately and may have enrollment periods that do not align with the Medicare open enrollment period.

Special Needs Plans (SNPs). In 2025, 1,445 SNPs will be offered nationwide, an 8 percent increase between 2024 and 2025 (Figure 2).

D-SNPs. Nearly two-thirds of SNPs (63%) are designed for people dually eligible for Medicare and Medicaid (D-SNPs). The number of D-SNPs has increased substantially since 2018, increasing from 401 D-SNPs in 2018 to 909 D-SNPs in 2025 (up from 851 D-SNPs in 2024), suggesting insurers continue to be drawn to this high-need population. In 2024, 5.9 million Medicare beneficiaries are enrolled in D-SNPs.

I-SNPs. The number of SNPs for people who require an institutional-level of care (I-SNPs) nearly doubled from 97 plans in 2018 to 189 plans in 2023, before dropping modestly to 173 plans for 2024 and 160 in 2025. In 2024, about 115,000 Medicare beneficiaries are enrolled in I-SNPs.

C-SNPs. The number of SNPs offered for people with chronic conditions (C-SNPs) has nearly tripled since 2018, from 132 plans that year to 376 plans in 2025 (an increase of 67 plans since 2024). Most C-SNPs focus on people with diabetes, heart disease, or lung conditions, as has been the case since the inception of C-SNPs. For 2025, one firm is offering a C-SNP for people with dementia (different than the one firm offering one in 2024). Two firms are offering C-SNPs for people with mental health conditions (compared to no firms in 2024), and one firm is offering a C-SNP for people with HIV/AIDS, the same one as in 2024. Ten firms are offering C-SNPs for people with end-stage renal disease (up two from 2024). In 2024, 675,000 Medicare beneficiaries are enrolled in C-SNPs.

Variation in Medicare Advantage Plans, by State and County. The number of plans available to the average beneficiary varies across states (Figure 3; Appendix Table 1). In 27 states and Puerto Rico, the average beneficiary has a choice of fewer plans in 2025 than in 2024. In 16 states and DC, the average beneficiary has access to more plans in 2025 than in 2024. In the remaining six states, the number of plans available to the average beneficiary stayed the same. This includes Alaska, which had no plans available in 2025, as in 2024. Connecticut is not included in this calculation – see methods for more details.

Nearly 2.2 million beneficiaries in 26 counties can choose from 75 or more Medicare Advantage plans in 2025 (down slightly from 29 counties in 2024). In the same 26 counties, beneficiaries can choose from 63 to 73 Medicare Advantage plans with Part D coverage. Similar to the last three years, the counties with the most plan options are predominantly in Pennsylvania and Ohio. In Pennsylvania, for example, beneficiaries can choose from 80 or more Medicare Advantage plans in 10 counties, including Dauphin County (Harrisburg). Beneficiaries in Cumberland, Pennsylvania can choose from 87 Medicare Advantage plans – the most offerings of any county in the U.S. In Ohio, beneficiaries can choose from 80 or more Medicare Advantage plans in 4 counties, including Summit County (Akron) and Cuyahoga County (Cleveland). In Michigan, beneficiaries in Oakland County (Detroit metro area) can choose from 75 plans (Figure 4).

In 2025, nearly one-third (32%) of Medicare beneficiaries (in 9 percent of counties) can choose from more than 50 Medicare Advantage plans (Figure 5).

Nearly one-third of Medicare beneficiaries (32%) have a choice of at least 50 Medicare Advantage plans in 2025, roughly the same as 2024 (33%). In 2025, less than 0.5 percent of beneficiaries live in a county with one to four Medicare Advantage plans available, while less than 0.5 percent of beneficiaries live in a county without any plans available. Similar to 2024, there are no Medicare Advantage plans for individual enrollment being offered in any county in Alaska in 2025, which includes about 99,000 beneficiaries. In 81 counties, about 250,000 Medicare beneficiaries (including those in Alaska) will not have access to a Medicare Advantage plan (an increase from 58 counties and about 196,000 Medicare beneficiaries in 2024). Additionally, no Medicare Advantage plans are available in territories other than Puerto Rico.

Variation in Medicare Advantage Plans by Geographic Status. Medicare beneficiaries living in metropolitan areas – counties with at least 50,000 people – can choose from 45 Medicare Advantage plans in 2025 on average (down from 47 in 2024), including 37 with Part D coverage, substantially more than beneficiaries living in rural or micropolitan areas. Beneficiaries in micropolitan areas (10,000 to 50,000 people) can choose from an average of 31 plans (1 fewer plan than in 2024), including 24 with Part D coverage.

Beneficiaries in rural areas – counties with less than 10,000 people – can choose from an average of 27 plans (the same as in 2024), including 21 with Part D coverage. Since 2020, when 14 plans were available on average to beneficiaries living in rural areas, the availability of plans in rural areas has nearly doubled.

As in recent years, virtually all Medicare beneficiaries (99.6%) have access to a Medicare Advantage plan as an alternative to traditional Medicare, including almost all beneficiaries in metropolitan areas (99.9%), micropolitan (99.0%) and rural (97.8%) areas.

Medicare Advantage Plan Availability by Firm

The average Medicare beneficiary is able to choose from plans offered by 8 firms in 2025, the same as in 2024 (Figure 6). Despite most beneficiaries having access to plans operated by several different firms, enrollment is concentrated in plans operated by UnitedHealthcare and Humana, and together UnitedHealthcare and Humana account for nearly half (47%) of Medicare Advantage enrollment in 2024.

In 2025, nearly one-third of beneficiaries (31%), in 169 counties, are able to choose from plans offered by 10 or more firms or other sponsors (a decline from 33% in 2024). In contrast, nearly 5 percent of beneficiaries live in a county where one to three firms offer Medicare Advantage plans (501 counties).

Further, in 126 counties, only one firm will offer Medicare Advantage plans in 2025. These are mostly rural counties with relatively few Medicare beneficiaries (less than 1 percent of total). In some of these counties, there were no firms offering Medicare Advantage in 2024, e.g., 5 counties in Idaho (Benewah, Clearwater, Custer, Lemhi, and Lewis). In contrast, Medicare beneficiaries in some counties had access to plans offered by two or three firms in 2024 but only one firm in 2025, such as people living in Coos County, Oregon.

Availability of Plans by Firm and County. UnitedHealthcare and Humana, the two firms with the most Medicare Advantage enrollees in 2024, have large footprints across the country, offering plans in most counties, similar to 2024 (Figure 7).

Some major insurers are expanding into new counties, while leaving others (Figure 8).

Humana is offering plans in more counties than any other large Medicare Advantage insurer – 2,848 counties in 2025 – though that represents a decrease of 58 counties from 2024. Humana is exiting 70 counties, while entering 12 new counties. UnitedHealthcare is offering plans in 2,808 counties in 2025, an increase of 4 counties from 2024. UnitedHealthcare entered more counties (42 new counties) than it exited (38 counties).

Blue Cross Blue Shield Affiliates are offering plans in 2,639 counties in 2025, an increase of 35 counties from 2024. Blue Cross Blue Shield Affiliates are exiting 44 counties, but are entering 79 new counties. CVS is offering plans in 2,249 counties, an increase of 22 counties. CVS exited 65 counties, but is entering 87 new counties.

Insurer decisions to exit some counties while entering new ones may suggest that local market factors impact insurer evaluations about their potential to attract enrollees and earn a profit.

Multiple Plan Offerings by Firms in the Same County. Many Medicare Advantage firms are also offering more than one plan option in each county.

More than half (58%) of Medicare beneficiaries live in a county (1,188) where at least one firm is offering 10 or more plans for individual enrollment. For example, in Cumberland County, Pennsylvania (the county with the most plan offerings – 87), four firms are offering 10 or more plans (Humana, Blue Cross Blue Shield Affiliates, UPMC Health Plan, and CVS Health). In 125 counties, two firms are offering 10 or more plans, and in 50 counties, three firms are offering 10 or more plans.

In 2025, Humana is increasing the number of plan options available in 631 counties, while decreasing the number of plan options in 1,192 counties. UnitedHealthcare is increasing the number of plans in 381 counties but decreasing plan offerings in 693 counties. Blue Cross Blue Shield Affiliates are increasing the number of plans options in 698 counties, while decreasing plan options in 630 counties. CVS Health is increasing the number of plan options in 667 counties, while decreasing them in 775 counties. This also suggests that insurers evaluate local markets when making decisions about the number of plans to offer and different county characteristics may make the market more or less attractive in a given year to a particular insurer.

Plan Renewals and Terminations

In 2025, 5% of Medicare Advantage enrollees in MA-PDs or about 1.4 million people, are in a plan that has been terminated for the coming year and will not be automatically assigned to another plan. (This number includes people enrolled in SNPs but excludes people in MA-only plans or people with employer coverage). People in this group will be able to enroll in another Medicare Advantage plan if one is available or choose traditional Medicare. If they choose traditional Medicare, they will qualify for a special enrollment period for Medigap with guaranteed issue rights, meaning they can switch to traditional Medicare and will not be denied a Medigap policy due to a pre-existing condition.

Another 7% of Medicare Advantage enrollees in MA-PDs or about 1.8 million people, are in plans that have been affected by a consolidation. In this situation, some portion of this 1.8 million people will be moved into another plan under the same insurer automatically if the contract includes another plan of the same type (i.e., HMO or PPO) in the same county. (Some enrollees in consolidated renewal plans will not see changes in their plan because they were already in the plan that other enrollees are now being assigned to.) They may still enroll in another Medicare Advantage plan if one is available or choose traditional Medicare. However, they do not qualify for a special enrollment period under federal law for Medigap.

New Market Entrants and Exits

In 2025, three firms (Healthy Mississippi, SECUR Health Plan and UCLA Health Medicare Advantage plan) entered the market for the first time, while eight firms exited the market (Appendix Table 2). Healthy Mississippi has one new HMO plan available for general enrollment in Mississippi, while UCLA Health is offering two HMO plans available for general enrollment in California. SECUR Health Plan is offering two I-SNPs in Florida.

In the last few years, some firms have introduced plans that are either co-branded or are in partnership with another company. For example, in 2025, Alignment Health is offering three plans co-branded with Instacart in 7 counties in California and Nevada. These plans will offer groceries to qualifying beneficiaries with chronic conditions. Alignment Health also partners with Walgreens and Rite Aid. Other companies with a partnership that are offering plans in 2025 include Select Health and Kroger and Humana and USAA though this is not an exhaustive list.

Eight firms that participated in the Medicare Advantage market in 2024 are not offering plans in 2025. Six of the firms had low enrollment in 2024 (around 10,000 or fewer enrollees per firm). Two of the firms had contracts taken over by other insurers.

Appendix

Note: This analysis was updated on November 25th to reflect the October version of the 2025 CMS Landscape file.

Medicare Advantage plans, which enrolled nearly 33 million Medicare beneficiaries or 54% of the eligible Medicare population in 2024, because they typically offer extra benefits, such as dental, vision and hearing, often for no additional premium, as well as lower cost sharing compared to traditional Medicare without supplemental insurance, with the trade-off of more restrictive provider networks and greater use of cost management tools, such as prior authorization.

This brief provides an overview of premiums and benefits in Medicare Advantage plans that are available for 2025 and key trends over time. This brief uses data from the CMS Landscape and Benefit files. See methods for more details. In general, this brief refers to Medicare plans available for general enrollment, excludes Special Needs Plans (SNPs), except where noted, and excludes employer plans. A companion analysis describes trends in plan offerings.

Medicare Advantage Highlights for 2025

• Two-thirds of all Medicare Advantage plans with Part D prescription drug coverage (MA-PDs) (67%) will charge no premium (other than the Part B premium) in 2025, similar to 2024 (66%).
• Nearly one-third (32%) of Medicare Advantage plans will offer some reduction in the Medicare Part B premium in 2025, an increase compared to 2024 (19%).
• Nearly all Medicare Advantage plans (97% or more) are offering vision, dental and hearing, as they have in previous years. However, the share of plans offering certain benefits has declined, such as over-the-counter benefits (85% in 2024 vs. 73% in 2025), remote access technologies (74% in 2024 vs. 53% in 2025), meal benefits (72% in 2024 vs. 65% in 2025) and transportation (36% in 2024 vs. 30% in 2025).
• The share of Special Needs Plans (SNPs) offering transportation, remote access technologies and in-home support services declined slightly in 2025, while the share offering bathroom safety devices and the Part B rebate increased.
• A larger share of SNPs than other Medicare Advantage plans are offering Special Supplemental Benefits for the Chronically Ill, which are extra benefits available to a subset of a plan’s enrollees, particularly food and produce (84% in SNPs; 15% in individual plans) and general supports for living, such as housing and utilities (67% in SNPs; 11% in individual plans).

Premiums

The vast majority of Medicare Advantage plans for individual enrollment (88%) will include prescription drug coverage (MA-PDs), similar to 2024 (89%), and the share of MA-PDs that charge no premium (other than the Part B premium of $185 per month) is 67% in 2025, similar to 2024 (66%). Nearly all beneficiaries (99%) have access to a MA-PD with no additional monthly premium in 2025, the same as in 2024 (99%).

In 2024, 75% of enrollees in MA-PD plans pay no premium other than the Medicare Part B premium. Based on enrollment in March 2024, 10% of enrollees pay at least $50 a month, including 3% who pay $100 or more. CMS estimates that the average monthly plan premium among all Medicare Advantage enrollees in 2025, including those who pay no premium for their Medicare Advantage plan, will be $17.00 a month. In 2024, 12 percent of Medicare Advantage enrollees are in a plan that offered some reduction in Medicare Part B premiums.

In 2025, 32% of Medicare Advantage plans will offer some reduction in the Part B premium, higher than the share in 2024 (19%) (Figure 1).

Among plans that are offering a monthly reduction in the Part B premium ($185 per month in 2025), 28% are offering a monthly reduction of $100 or more, 25% are offering a reduction of $50.01 to $100, 17% are offering a reduction of $10.01 to $50, and 30% are offering a monthly reduction of $10 or less.

In previous years, a smaller share of Medicare Advantage enrollees has typically ended up in plans that reduced the Part B premium. For example, for the 2024 plan year, 19% of plans offered a reduction in the Part B premium, but ultimately only 12% of Medicare Advantage enrollees were enrolled in plans with this benefit.

While many employers and unions also offer Medicare Advantage plans to their retirees, no information about these 2025 plan offerings is made available by CMS to the public during the Medicare open enrollment period. Employer and union plans are administered separately and may have enrollment periods that do not align with the Medicare open enrollment period.

Extra Benefits

Medicare Advantage plans may provide extra benefits that are not available in traditional Medicare, are considered “primarily health related,” and can use rebate dollars (including bonus payments) to help cover the cost of these extra benefits. Beginning in 2019, CMS expanded the definition of “primarily health related” to allow Medicare Advantage plans to offer additional supplemental benefits. Medicare Advantage plans may also restrict the availability of these extra benefits to certain subgroups of beneficiaries, such as those with diabetes or congestive heart failure, making different benefits available to different enrollees.

Availability of Extra Benefits in Individual Plans for General Enrollment. In 2025, 97% or more individual plans offer some vision, dental or hearing benefits, similar to 2024 (Figure 2). Though these benefits are widely available, the scope of coverage for these services varies. For example, a dental benefit may include cleanings and preventive care or more comprehensive coverage, and often is subject to an annual dollar cap on the amount covered by the plan. From year to year, plans may change the parameters of this coverage, such as increasing or decreasing annual maximums the plan will pay toward the benefit or adjusting cost sharing for services. There is not yet data available about utilization of these benefits or associated costs, so it is not clear the extent to which supplemental benefits are used by enrollees.

As of 2020, Medicare Advantage plans have been allowed to include telehealth benefits as part of the basic benefit package – beyond what was allowed under traditional Medicare prior to the COVID-19 public health emergency, which was extended to December 2024. Therefore, these benefits are not included in the figure above because their cost is not covered by either rebates or supplemental premiums. Medicare Advantage plans may also offer supplemental telehealth benefits via remote access technologies and/or telemonitoring services, which can be used for those services that do not meet the requirements for coverage under traditional Medicare or the requirements for the telehealth benefits as part of the basic benefit package (such as the requirement of being covered by Medicare Part B when provided in-person). In 2025, 53% of plans are offering remote access technologies, a decline from 74% in 2024. A similar share of plans are offering telemonitoring services (2% in 2025 vs 3% in 2024).

Some benefits are being offered by a smaller share of plans in 2025 than in 2024. For example, 73% of plans are offering an allowance for over-the-counter items (vs. 85% in 2024), while 65% are offering meal benefits (vs. 72% in 2024), and 30% are offering transportation benefits for medical needs (vs. 36% in 2024). A similar share of plans is offering acupuncture (32% in 2025 vs. 34% in 2024), bathroom safety devices (24% in 2025 vs 22% in 2024), and support for caregivers of enrollees (5% in 2025 and 2024). A smaller share of plans are offering in-home support services (6% in 2025 vs 9% in 2024). This is not an exhaustive list of extra benefits that plans offer, and plans may provide other services such as home-based palliative care, therapeutic massage, and adult day health services, among others.

Access to Medicare Advantage Plans with Extra Benefits. Virtually all Medicare beneficiaries live in a county where at least one Medicare Advantage plan available for general enrollment (excluding SNPs) has some extra benefits not covered by traditional Medicare, with over 99% having access to at least one or more plans with dental, fitness, vision, and hearing benefits for 2025, the same as in 2024. The vast majority of beneficiaries also have access to one or more plans that offer an allowance for over-the-counter items (over 99%), a meal benefit (over 99%), remote access technologies (99%), acupuncture (98%), bathroom safety devices (96%), transportation assistance (94%) but fewer have access to one or more plans that offer in-home support services (60%), caregiver support (41%), or telemonitoring services (16%).  (Connecticut is not included in these estimates – see methods for more details.)

Availability of Extra Benefits in Special Needs Plans. SNPs are designed to serve a disproportionately high-need population, and a somewhat larger percentage of SNPs than plans for other Medicare beneficiaries provide their enrollees over-the-counter benefits (92%; similar to 2024 – 94%), transportation benefits for medical needs (81%; a decline from 88% in 2024), meals (73%, similar to 2024 – 75%), bathroom safety devices (54%; up from 34% in 2024), and in-home support services (17%; down from 25% in 2024) (Figure 3). Compared to individual plans, a smaller share of SNPs offer fitness benefits (83%, similar to 2024 – 84%), remote access technologies (50%; a decline from 66% in 2024), and the Part B rebate (29%; up from 7% in 2024). Similar to plans available for individual enrollment, a relatively small share of SNPs offer support for caregivers (5%) or telemonitoring services (2%).

Availability of Special Supplemental Benefits for the Chronically Ill (SSBCI). Beginning in 2020, Medicare Advantage plans have also been able to offer extra benefits to a subset of a plan’s enrollees, that are not primarily health related and are specifically for chronically ill beneficiaries, known as Special Supplemental Benefits for the Chronically Ill (SSBCI). In addition, Medicare Advantage plans participating in the Value-Based Insurance Design Model may also offer these non-primarily health related supplemental benefits to their enrollees, but can use different eligibility criteria than required for SSBCI, including offering them based on an enrollee’s socioeconomic status (e.g., LIS eligibility) or whether the enrollee lives in an underserved area.

Most individual and SNP Medicare Advantage plans still do not offer these benefits, though more SNP plans generally offer these benefits, particularly food and produce. SSBCI benefits offered in 2025 include food and produce (15%% for individual plans and 84% for SNPs), general supports for living (e.g., housing, utilities) (11% in individual plans and 67% for SNPs), transportation for non-medical needs (8% for individual plans and 46% for SNPs), and pest control (3% for individual plans and 23% for SNPs) (Figure 4).

Like for other types of supplemental benefits, the scope of services for SSBCI benefits varies. For example, many plans offer a specified dollar amount that enrollees can use toward a variety of benefits, such as food and produce, utility bills, rent assistance, and transportation for non-medical needs, among others. This dollar amount is often loaded onto a flex card or spending card that can be used at participating stores and retailers, which can vary depending on the vendor administering the benefit. Depending on the plan, this may be a monthly allowance that expires at the end of each month or rolls over month to month until the end of the year, when any unused amount expires.

With Donald Trump returning to the presidency, the future of Medicaid is uncertain. While Medicaid did not receive a lot of attention directly during the campaign, Trump’s first term can shed light on potential changes that could be implemented administratively without Congress.

1. Trump administration could encourage and approve Medicaid waivers to advance priorities, including work requirements.

The previous Trump administration’s Section 1115 waiver policy emphasized work requirements, other eligibility restrictions, and capped financing. The Trump administration approved 13 waivers that allowed states to condition Medicaid eligibility on meeting work and reporting requirements and also approved waivers that restricted eligibility, including by permitting states to charge premiums and lock out enrollees who are disenrolled for unpaid premiums. Although previous work requirements in Medicaid were challenged in court and the waivers imposing work requirements were rescinded by the Biden administration, a number of states continue to be interested in adopting them. Although CMS withdrew work and premium requirements in Georgia’s “Pathways” waiver, these provisions remain in place after a federal judge vacated the CMS rescission.

Trump also introduced a demonstration opportunity, the “Healthy Adult Opportunity” (HAO), that would have allowed states “extensive flexibility” to use Medicaid funds to cover certain adults – including those who qualify under the Affordable Care Act’s expansion — without being bound by federal standards related to eligibility, benefits, delivery systems, and oversight in exchange for annual limits on federal financing. While no state took up this option during the first administration, these waivers restructure Medicaid financing in ways similar to block grant and per capita cap proposals and promoting the waivers in a second administration would enable Trump to advance that policy even in the absence of Congressional action. However, unlike legislative proposals that would alter Medicaid financing and significantly cut federal spending, it is unlikely that states would opt for such deals if federal financing would be cut significantly.

The new Trump administration’s waiver priorities will differ significantly from those of the Biden administration and it is unclear how the Trump administration will treat certain waivers promoted and approved by the Biden administration, such as those focused on addressing health-related social needs, multi-year continuous eligibility, and leveraging Medicaid to help individuals leaving incarceration transition to the community. The Trump administration could choose not to approve waivers that remain pending, rescind existing waiver guidance, or withdraw approved waivers, although some of these waivers, particularly ones using Medicaid to assist with reentry from incarceration, have been pursued by both Republican and Democratic governors.

2. Trump administration could choose not to implement or rewrite recent regulations.

The Biden administration finalized a number of major Medicaid regulations designed to promote quality of care and advance access to care for Medicaid enrollees as well as to streamline eligibility and enrollment processes in Medicaid and the Children’s Health Insurance Program (CHIP).

• The Access rule addresses several dimensions of access: increasing provider rate transparency and accountability, standardizing data and monitoring, and creating opportunities for states to promote active enrollee engagement in their Medicaid programs. The rule also included many provisions governing access to home care (also known as home- and community-based services or HCBS), which include ensuring that at least 80% of spending on certain services be spent on compensation for direct care workers and requiring states to report the number of people on waiting lists for care.
• The Managed Care rule addresses Medicaid managed care access, financing, and quality, including strengthening standards for timely access to care and states’ monitoring and enforcement efforts.
• The Long-Term Care Facility (LTC) Staffing rule requires minimum staffing standards for nursing facilities.
• Two rules streamline Medicaid enrollment and renewal processes for the Medicare Savings Program (MSP) and for Medicaid, CHIP and the Basic Health Program. The first rule helps eligible Medicare beneficiaries more easily access Medicaid coverage of Medicare premiums and cost sharing through the MSP while the second rule simplifies application, enrollment, and renewal processes and removes access barriers for children in CHIP. Each rule is expected to increase Medicaid enrollment by about one million people.

These rules are complex and are set to be implemented over several years. The Trump administration could delay implementation of certain provisions, which would reduce regulation of managed care companies, nursing facilities, and other providers, while rolling back enrollee protections, payment transparency, and improved access. Alternatively, the Trump administration could issue new regulations that would undo these final regulations.

3. Trump administration could issue guidance and implement policy to make it more difficult for people to obtain and maintain coverage, reducing enrollment and spending.

The Trump administration could also issue sub-regulatory guidance to change eligibility or renewal policies to restrict coverage by tightening verification of eligibility. Previously, the Trump administration sought to reduce Medicaid enrollment by encouraging states to conduct eligibility verification processes in between annual renewal periods. While some states conduct these periodic data checks, the administration could encourage more states to adopt the policy as a program integrity strategy. The Trump administration could also eliminate flexibilities granted under the Biden administration that allow states to streamline renewal procedures. Getting rid of these flexibilities would likely make it harder for some enrollees to renew their Medicaid coverage, while providing greater certainty that people who are ineligible do not obtain coverage.

Beyond waivers, regulations, and enrollment policy, if the Republicans gain control of both the House and the Senate, Trump could work with Congress to enact legislation that would more fundamentally change the financing structure of Medicaid and make significant cuts to federal Medicaid spending, similar to policies Trump has supported in the past.

Abortion drove many voters to turn out for Tuesday’s election, but not always for Vice President Kamala Harris, while concerns about the economy weighed more heavily on voters’ minds, according to polling data from KFF and the Associated Press.

About a quarter of voters said abortion was the “single most important” factor in their vote, about 4 in10 said it had a major impact on their decision to turn out (similar to the shares who said so in the 2022 midterm elections), and over half said it had a major impact on which candidates they supported, according to the Associated Press’s VoteCast survey. In partnership with the AP, KFF added supplemental questions to the survey of the 2024 electorate to gauge how voters were weighing health care issues as they exited the polls.

But Vice President Harris’s strong advantage on abortion was not enough to override negative views of the economy and immigration, issues where President- Elect Donald Trump held the edge, the polling data shows. Forty percent of voters said high prices for gas, groceries and other goods was the single most important factor to their vote, including 51% of Republicans, 41% of independents, and 28% of Democrats.

KFF’s analysis shows that Trump did much better than Harris among the majorities of voters who reported being very concerned about the cost of groceries, gas, and housing. He also did better than Harris among the more than half of voters who said they were very concerned about the cost of health care.

Moreover, on the issue of abortion, Trump managed to garner small but important shares of votes among those who voted in favor of abortion ballot measures in their state, winning four states where voters also chose to expand or protect abortion access. Trump won support from about three in ten voters who voted in favor of abortion access in the battleground states of Nevada and Arizona, a similar share in Missouri, and about a quarter in Montana.

In an online interactive dashboard, KFF examines the role that abortion, the economy and health care costs played in the 2024 election, providing a deep dive into voters’ thinking on health care issues as they made their decisions about whether to vote and whom to vote for.

KFF provides data and analysis of the overall AP VoteCast survey and these supplemental questions at both the national and state level.

The dashboard includes:

• An interactive map that provides a more detailed look at states with abortion-related ballot measures.
• Analysis and charts examining the motivations of voters and the impact of concerns about abortion policy and health care costs nationally and across states, including some data breakouts by demographic groups and political party.
• A curated set of charts providing key analysis among subgroups of voters on topics like abortion ballot measures, health care as an economic concern, and voters’ views of the most important issues facing the U.S.

The AP VoteCast is a national and 48 state surveys of more than 115,000 voters conducted by NORC at the University of Chicago for The Associated Press, Fox News, PBS NewsHour, and The Wall Street Journal beginning on Oct. 28 and concluding as polls closed on Nov. 5, 2024, in English and Spanish. The national survey was conducted using NORC’s probability-based AmeriSpeak panel, while the individual state surveys were conducted from a random sample of state voter files and from self-identified registered voters selected from non-probability online panels. More details are available about AP VoteCast’s methodology here.

GLP-1 (glucagon-like peptide-1) drugs have been used as a treatment for type 2 diabetes for over a decade, but newer forms of these drugs have gained widespread attention for their effectiveness as a treatment for obesity. While these drugs have provided new opportunities for obesity treatment, they have also raised questions about access to and affordability of these drugs. These drugs are expensive when purchased out of pocket, and coverage in Medicaid, ACA Marketplace plans, and most large employer firms remains limited, though a number of state Medicaid programs and other payers are re-evaluating their coverage policies. Expanding Medicaid coverage of these drugs could increase access for the almost 40% of adults and 26% of children with obesity in Medicaid. At the same time, expanded coverage could also increase Medicaid drug spending and put pressure on overall state budgets. In the longer term, however, reduced obesity rates among Medicaid enrollees could also result in reduced Medicaid spending on chronic diseases associated with obesity, such as heart disease, type 2 diabetes, and types of cancer. This brief discusses Medicaid coverage of GLP-1s, examines recent trends in Medicaid prescriptions and gross spending on GLP-1s, and explores the potential implications of expanding coverage obesity drugs for Medicaid programs.

Does Medicaid cover GLP-1s for obesity treatment?

States can decide whether to cover obesity drugs under Medicaid. Under the Medicaid Drug Rebate Program, Medicaid programs must cover nearly all of a participating manufacturer’s Food and Drug Administration (FDA)-approved drugs for medically accepted indications. However, weight-loss drugs are included in a small group of drugs that can be excluded from coverage¹ . Though the statutory exception refers to agents used for “weight loss”, “obesity drugs” is used to refer to this group of medications in this analysis. The FDA has approved three GLP-1s for the treatment of obesity, Saxenda (liraglutide), Wegovy (semaglutide), and Zepbound (tirzepatide), and state Medicaid coverage of these is optional. However, Medicaid programs have to cover formulations to treat type 2 diabetes, including Ozempic (semaglutide), Rybelsus (semaglutide), Victoza (lirglutide), and Mounjaro (tirzepatide). Wegovy, as of March 2024, must also be covered for preventing heart attacks or strokes in adults with cardiovascular disease; however, this expanded label indication does not impact this analysis as it only includes data through 2023. Notably, all obesity drugs are covered for children under Medicaid’s Early and Periodic Screening, Diagnostic and Treatment (EPSDT) benefit, though it is less clear how states are implementing and covering in practice.

Obesity drug coverage in Medicaid remains limited, with 13 state Medicaid programs covering GLP-1s for obesity treatment as of August 2024 (Figure 1). Twelve states in KFF’s annual budget survey reported coverage of GLP-1s for obesity treatment under FFS as of July 1, 2024, and North Carolina reported adding coverage in August of 2024. All 12 states that reported coverage of GLP-1s as of July 1, 2024 also reported that utilization control(s) applied, with the most common being prior authorization (11 of 12 states) and/or BMI requirements (11 of 12 states). Eleven of the 12 states reported covering all three GLP-1s currently approved for the treatment of obesity (Saxenda, Wegovy, or Zepbound). While the survey only asked about FFS coverage, MCO drug coverage must be consistent with the amount, duration, and scope of FFS coverage. MCOs, however, may apply differing utilization controls and medical necessity criteria unless the state’s MCO contract specifies otherwise. Coverage among other payers also remains limited. Recent KFF analysis also found most large employer firms do not cover GLP-1 drugs for weight loss, coverage in ACA Marketplace plans remains limited, and coverage in Medicare is prohibited.

How have Medicaid prescriptions and gross spending on GLP-1s changed in recent years?

The number of Medicaid prescriptions and gross spending on GLP-1s have increased rapidly in recent years, with both nearly doubling from 2022 to 2023. Overall, from 2019 to 2023, the number of GLP-1 prescriptions increased by more than 400%, while gross spending increased by over 500%. Spending per prescription before rebates reached more than $900 per prescription in 2023. Those prices and spending numbers do not account for rebates, and states are likely receiving substantial rebates on these brand drugs. While rebate data for specific drugs is not publicly available, Medicaid and CHIP Payment and Access Commission (MACPAC) analysis of FY 2020 data found statutory rebates accounted for 61.6% of gross Medicaid spending on brand drugs. Also, amid growing criticism of the cost of their drugs, Novo Nordisk, the company that creates Ozempic and Wegovy, has said that rebates and other fees (across all payers) account for about 40% of the cost of the two drugs. The GLP-1s in this analysis still account for relatively small shares of the total number of Medicaid prescriptions and spending before rebates, though the shares are growing. By 2023, these drugs accounted for 0.5% of all Medicaid prescriptions (up from 0.01% in 2019) and 3.7% of all gross Medicaid spending (up from 0.9% in 2019).

Specifically, increased utilization of Ozempic, Wegovy, and Mounjaro have contributed substantially to recent growth. Prescriptions and spending on Ozempic, approved to help control blood sugar levels for adults with type 2 diabetes in 2017 (Table 1), have grown considerably from 2019 to 2023, nearly doubling every year since 2019. Looking from 2022 to 2023, the latest year of data available, Wegovy (first approved in 2021) and Mounjaro (approved in 2022) also saw substantial growth, with prescriptions and gross spending for both drugs increasing twelvefold or more. This analysis includes GLP-1 formulations approved for obesity treatment as well as the same formulations approved for type 2 diabetes (for more information on how GLP-1s are identified in this analysis, see Methods). From Medicaid data publicly available, there is no way yet to disentangle how much of the growing use of GLP-1s is related to treatment for diabetes versus obesity, or a combination of both. In addition, the popularity and increased demand for GLP-1s has led to drug shortages, sometimes causing people to switch products or ration doses or sometimes leaving individuals without access to needed prescriptions. This may have implications for drug trends, though the FDA recently reported that GLP-1 supplies are beginning to stabilize.

How may the coverage landscape of GLP-1s for obesity treatment change?

Many state Medicaid programs are considering covering obesity drugs in the future but are concerned about the cost implications. KFF’s annual budget survey found that, among those states that do not currently cover obesity drugs, half reported they were considering adding coverage, with a few states reporting plans to add or expand coverage in FY 2025 or later. When asked about the key factors contributing to their obesity medication coverage decision, almost two-thirds of responding states mentioned cost, though states are also weighing a number of other factors including the need for legislative action, adherence concerns, clinical criteria development, and potential side effects. Conversely, 4 in 10 states noted that positive health outcomes and longer-term savings on chronic diseases associated with obesity were key factors in their decision to cover or consider covering in the future along with increasing enrollee access and health equity, recommendations from providers, and ability to negotiate supplemental rebate agreements. States are likely considering various cost containment strategies for these drugs and may even be re-evaluating their broader approach to obesity treatment, including the use of obesity medications along with other treatments such as nutritional counseling or behavioral therapy. Obesity is caused by a multitude of complex factors, and uptake of GLP-1s could improve health but would not address all of the underlying contributors to obesity.

1. Drugs that may be excluded from coverage under the MDRP include drugs used for: a) anorexia, weight loss, or weight gain, b) promoting fertility, c) cosmetic purposes or hair growth, d) symptomatic relief of cough and colds, e) prescription vitamins and mineral products except prenatal vitamins and fluoride preparations , f) nonprescription drugs, g) a manufacturer’s covered outpatient drug in which tests and monitoring services have to be purchased from that manufacturer, h) sexual or erectile dysfunction unless used to treat a condition. For more information see, 42 U.S.C. § 1396r-8. ↩︎

Introduction

How we ask for, analyze, and report information on race and ethnicity affects our ability to understand the racial and ethnic composition of our nation’s population and our ability to identify and address racial disparities in health and health care. The accuracy and precision of such data have important implications for identifying needs and directing resources and efforts to address those needs. Race, ethnicity, and national origin are distinct concepts that are social constructs, and how they have been defined, identified, and/or categorized have evolved over time. This brief provides an overview of how the concepts of race, ethnicity, and nationality have been defined and measured by the federal government through the U.S. Census Bureau and the Office of Management and Budget (OMB) over time and the implications for health disparities. We acknowledge that this brief does not cover all the nuances and complexities of the topic of racial and ethnic identity and that there is variation in how people think, talk, and relate to race, ethnicity, and national identity.

What do Race, Ethnicity, and National Origin Represent?

The concepts of race, ethnicity, and national origin and their fluidity are reflective of these identities being social constructs. While different, the two concepts of race and ethnicity are connected. Race is defined as a social political category primarily based on physical characteristics such as skin color, and ethnicity is a social category primarily defined by culture, language, and history. National origin is defined by the country or region that an individual or their ancestors originate from. In a 2024 update, federal standards for collecting and reporting racial and ethnic data combined previously separate questions about race and Hispanic ethnicity into a single question, and a new category was added for Middle Eastern or North African (MENA) people. The updated federal standards utilize seven racial and ethnic categories that are identified with the following terms (see Appendix for more details):

• American Indian or Alaska Native (AIAN)
• Asian
• Black or African American
• Hispanic or Latino
• Middle Eastern or North African (MENA)
• Native Hawaiian or Pacific Islander (NHPI)
• White

There are distinctions in the meanings of and community preferences for different terms, such as Black and African American and Hispanic, Latino, or Latinx. Polling data over time show that half of those who trace their roots to Spanish-speaking Latin America and Spain have consistently said they have no preference for Hispanic or Latino, but that when asked to choose one term over another, Hispanic has been preferred to Latino. Surveys further show a preference for country-of-origin labels (such as Mexican, Cuban, or Ecuadorian) versus broader pan-ethnic terms. A 2023 survey of U.S. Hispanic adults found that 47% of U.S. adults who self-identify as Hispanic have heard of the term Latinx, and just 4% say they use it to describe themselves. A 2021 Gallup Poll similarly found that most people favor the use of the term Hispanic and few (4%) prefer Latinx. Polling data also show that most Black Americans do not have a preference between Black and African American when asked which term they would rather people use to describe their racial group.

A large and growing share of people identify with more than one of the previous federal racial and ethnic categories (which do not reflect the 2024 changes to the standards). KFF analysis of 2023 American Community Survey (ACS) data finds that about eight in ten (81%) AIAN and two thirds (66%) of NHPI people identify with more than one racial and/or ethnic group. About one in five Asian (20%), Black (18%), and White (21%) people also identify with more than one racial and/or ethnic group (Figure 1). Under previous standards which had separate questions for Hispanic ethnicity and race, people of Hispanic ethnicity may be of any race.

Racial identity is important to how people think about themselves, particularly for Black and Hispanic people. In a recent KFF survey, a majority of Black (83%) and Hispanic (70%) people said that their racial identity is very or extremely important to how they think about themselves. About half of Asian (51%) people say the same. Data were unavailable for other racial and ethnic groups (Figure 2).

How Have Measures of Race, Ethnicity, and National Origin Evolved Over Time?

Since its inception in 1790, the U.S. census has collected information on race and ethnicity that has informed policy, the allocation of resources, and scientific research on different groups within the country. The race and ethnicity data collected by the census is also used to evaluate the effectiveness of government programs and policies as well as to measure and ensure fairness, equity, and compliance with anti-discrimination laws and policies. In 1977, OMB established federal standards for race and ethnicity data through Statistical Policy Directive No. 15 to standardize how data are collected and reported at the federal level. The OMB standards guide not only the census but also other federal surveys, ensuring consistency across government data collection efforts. The census aligns its racial and ethnic categories with these standards, though it occasionally adjusts them to reflect evolving understandings of identity. The census recognizes that race is a social construct, stating that “The racial categories included in the census questionnaire generally reflect a social definition of race recognized in this country and not an attempt to define race biologically, anthropologically, or genetically. In addition, it is recognized that the categories of the race item include racial and national origin or sociocultural groups.”

How the census has collected and categorized information on race and ethnicity has evolved significantly over time, reflecting social and political shifts and the growing diversity of the U.S. population (Figure 3).

• Between 1790 and the mid-20^th century, race and ethnicity information was collected via enumerators who conducted an interview for each household. Enumerators identified individuals’ racial and ethnic identities through their observations based on criteria that included physical characteristics, social norms and principles, national origin, and Tribal affiliation. Between 1790 and 1860, the census collected information on White, Black (enslaved and free), and Other Free Persons.¹ 
• Additional categories were added to the census in the 1800s, which were motivated by slavery and race science and maintaining rights and privileges for White people. The 1840 census introduced a new category, free colored people (including Black and mixed-race Black people). In the mid to late 1800s, mixed-race categories were added — “mulatto, quadroon, and octoroon”. The definitions of these categories evolved over time, but they were based on perceived shares of Black ancestry and are no longer used. American Indian people who renounced Tribal rule, exercised the rights of U.S. citizens, and paid taxes were counted in the census for the first time in 1860. Prior to that, the Constitution excluded American Indian people who lived on reservations, lived on unsettled land, or were not taxed from being enumerated. By the late 1800s, the census began collecting national origin information, coinciding with the entry of Chinese migrant workers.
• Amid the rise of Jim Crow laws, the 1930 census dropped mixed-race categories and focused on the collection of single race data. Up until 1970, census enumerators classified people of mixed-race heritage based on specific rules: those with White ancestry and another racial group were classified as “non-White,” and people with two “non-White” ancestries were classified by the father’s race. If a person was Black or AIAN, their racial identity was instead based on blood quantum rules. Blood quantum measures the percentage of American Indian or Alaska Native ancestry or blood that an individual has and is used to determine Tribal affiliation. People who were both Black and White were categorized as Black under the one-drop rule, which meant that a person with any percentage of Black ancestry or blood would be counted as Black.
• Also during this period, more disaggregated Asian national origin data were collected, and Mexican national origin data were collected for the first time. Before 1930, Mexican Americans were classified as White. There was an organized movement to remove the Mexican category and reclassify Mexican people as White. Mexican people were reclassified as White in the 1940 census until 1970 when the census added the Hispanic/Latin origin category to a version of the questionnaire that was sent to a small share of the U.S. population.
• In the mid-20^th century, the census shifted to allow respondents to self-identify their own race and/or ethnicity. Respondents were instructed to select the race that they most closely identified with from the single-race categories available or to use the father’s race if they were uncertain. Further, the addition of two new states in 1959, Alaska and Hawaii, prompted the addition of new categories to the census in 1960; Eskimo, Aleut, Hawaiian, and part-Hawaiian.
• In 1980, the census added a separate question on Hispanic ethnicity, following lobbying efforts from Hispanic advocacy groups in the 1970s in response to undercounting of the Hispanic population in the census. The Asian Pacific Islander (API) category was also added at this time in response to lobbying by Asian American legislators and advocacy groups, who similarly found that Asian American and Pacific Islander people were being undercounted. The API indicator included nine detailed Asian and Pacific Islander origin categories.
• In 1997, the OMB revised its race and ethnicity standards to allow individuals to select more than one racial category, reflecting a growing recognition of multiracial identities. This revision aimed to improve data accuracy and better capture the diversity of the U.S. population. Other open-ended questions on ancestry or ethnic origin were added that led to broader ethnic identification among American Indian people with a large share of the population claiming some Indian ancestry even if they didn’t identify racially as American Indian. This may have reflected increased American Indian pride movements that prompted people with multiracial American Indian heritage to identify with their American Indian ancestry.
• The second major change that impacted people with mixed-race heritage occurred in 2000, when the census allowed people to select more than one racial category. This change was driven by the multiracial movement of the 1980s and 1990s. In 2000, the census combined the American Indian and Alaska Native categories to form the AIAN category to capture original peoples with origins in North, Central, and/or South America. It also asked respondents to provide the names of their enrolled or principal Tribes. Additionally, it began collecting information separately on NHPI people, which it defined as people having origins in any of the original peoples of Hawaii, Guam, Samoa, or other Pacific Islands.
• The 2020 census allowed people to include detailed information about their race and/or ethnicity in addition to marking multiple race categories. The census now allows respondents to provide write-in responses describing their race and ethnicity, with clearer instructions and examples based on the largest population groups for each category. The multiracial population grew significantly between 2010 and 2020, with the U.S. Census Bureau indicating that changes in the design, data processing, and coding of the race and ethnicity questions over this time period (including the write-in responses) contributed to this growth, highlighting the impact of these decisions. This change also led to more people of MENA heritage providing detailed information about their ancestry, whereas before they were aggregated into the White category.
• In 2024, OMB released revised Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity to better reflect the growing diversity of the U.S. population. The revisions include using a single combined question for race and ethnicity, adding MENA as a minimum category, clarifying instructions for individuals to select multiple racial and ethnic categories that represent their identity, and requiring collection of more detail beyond the minimum categories. In addition, the standards require that data tabulation procedures result in the production of as much information on race and/or ethnicity as possible, including data for people reporting multiple racial and/or ethnic categories. These changes will impact how race and ethnicity data are collected in the census, as the U.S. Census Bureau has to adhere to the OMB standards on race and ethnicity.

While the census and OMB standards have been the standard for measuring race and ethnicity in the U.S. government at the federal level, states, localities, and other organizations often differ in their measurement of racial and ethnic data. For example, states vary in the number of categories they use to collect race and ethnicity data, as well as how these variables are named and combined. In Oregon, as part of an effort to eliminate health inequities, Oregon’s Health Authority has taken steps to accurately and expansively report demographic data. This includes collecting detailed and more granular race, ethnicity, language, and disability data. Oregon collects data for 42 race and ethnicity groups, answers are self-reported, and respondents are given the opportunity to select more than one race and/or ethnicity. Some states have narrower race and ethnicity categories, combining groups such as Asian and Native Hawaiian or Pacific Islanders into one group, API.

How Do Measures of Race, Ethnicity, and National Origin Impact Policies and Health Disparities?

How race, ethnicity, and nationality have been defined and measured has important implications for health disparities. Historically, these measures affected who can access health care, social services, education, and employment opportunities and have reinforced racial misinformation that limit access to resources for marginalized groups. Narrow and inconsistent race and ethnicity categories have obscured inequities and made it more difficult to meet the diverse needs of different populations. Conversely, data on race and ethnicity have also been used to address disparities by informing policies and interventions and to ensure compliance with antidiscrimination laws.

There are historic examples of racial and ethnic data being used in ways that have worsened and perpetuated racism and disparities. For example, in 1840, an “insane or idiot” category was added to the census to identify the number of people with mental disabilities in the country. However, the census enumerators disproportionately overcounted free colored people as “insane” to support the inaccurate idea that “freedom drove Black people mad.” Advocates of slavery used the 1840 census data to justify that slavery was beneficial for the health and well-being of Black people. In 1850, scientists petitioned the addition of a new racial category, “Mulatto” (people with mixed Black ancestry),  to study the health of multiracial enslaved people. Census data from this period contributed greatly to scientific racism. Scientists used this data to suggest people of color were inferior to White people, test theories of polygenism, and codify racial hierarchies. As the multiracial population grew, so did ideologies surrounding the rules of hypodescent, under which multiracial individuals are assigned the race of the parent from the marginalized racial group. This included the one-drop rule, which required that anyone with a discernible trace of African ancestry be considered Black. This concept ensured that the children of enslaved Black people and their White enslavers would remain slaves. Blood quantum categorization eventually led to the codification of the one-drop rule in some states during the Jim Crow era as a means of supporting segregation. Similarly, the census began collecting information on American Indian blood quantum in 1930. Blood quantum was not only used to determine Tribal membership but was also used to study any perceived biological and intellectual differences between American Indian and White people.

In addition, race has historically and continues to play a role in medical teaching and clinical decision making within health care. Historically, the medical and scientific community used race to explain differences in disease prevalence and outcomes, contributing to misperceptions about biological differences by race that were used to justify mistreatment. Within U.S. medical curricula, the concept of race led to since disproven theories of biological inferiority of people of color and White supremacy, which fueled an array of atrocities in medicine including the forced sterilization efforts targeting Black and Native American women, the use of Henrietta Lacks’ cells for scientific research without consent, and the infamous U.S. Public Health Service Untreated Syphilis Study at Tuskegee, among others. Today, research suggests that provider and institutional bias and discrimination are drivers of disparities and health. Race also continues to be used as a factor in some clinical algorithms, although there is growing movement to eliminate the use of race and to ensure that disparities are not perpetuated amid the growing use of artificial intelligence and algorithms to guide clinical decision making.

Conversely, racial and ethnic measures have also been used to mitigate inequities in policies, employment, health care, and other sectors. For example, the Civil Rights Movement in the mid-20^th century prompted the standardization of racial and ethnic classification as well as documentation of trends in racial and ethnic discrimination. This resulted in the establishment of the OMB Statistical Policy Directive No. 15 in 1977 that has since standardized the collection of race and ethnicity data at the federal level. The collection of standardized data facilitates the ability for policymakers and institutions to identify and address areas of inequality. For instance, racial and ethnic data can inform resource allocation, ensuring communities facing systemic disadvantages receive essential services, such as health care, education, and social programs. Additionally, it can help track and mitigate racial bias and discriminatory practices in health care, employment, housing, and other social and economic domains. These data also allow for the evaluation of the effectiveness of interventions designed to reduce disparities. For example, public health programs can measure the impact of vaccinations, screenings, or outreach efforts in marginalized communities, using racial and ethnic data to refine strategies and ensure more equitable access to care.

Availability of racial and ethnic data also has impacts on efforts to address health disparities. Missing or inconsistent data on race, ethnicity, and nationality can hinder effective resource allocation and policy decision-making, particularly in efforts to address health disparities. AIAN people were excluded in early versions of the census, beginning a trend of exclusion from national data inquiry that continues to the present day. This exclusion from data and analysis has contributed to limiting the visibility and understanding of challenges faced by AIAN people and other smaller racial and ethnic groups, including NHPI people. The negative impacts of missing or incomplete data were evidenced during the COVID-19 pandemic, when inconsistencies and limitations in how states reported their data limited the ability to understand racial and ethnic disparities in COVID-19 health impacts as well as take-up of COVID-19 vaccinations. Arab Americans and people with ancestry in the Middle East or North Africa have been invisible in key datasets, resulting in a limited understanding of their health outcomes, experiences accessing health care, and engagement with the health care system. Increasing the availability of disaggregated racial and ethnic data facilitates a greater understanding of disparities in health and health care and can help focus efforts to address them. For example, while aggregate data on Asian people suggest that they fare the same or better as compared to White people across most measures of health and health care, they mask underlying disparities among smaller subgroups within the Asian community. Having more disaggregated data allows for a more nuanced understanding of people’s experiences and can facilitate focused efforts to address disparities as well as to measure impacts of interventions to address them.

What are Future Issues to Consider Related to Measuring Race, Identity, and National Origin and Health Disparities?

As the U.S. becomes more diverse, it will be increasingly important to consider how to identify people’s identities, particularly among multiracial people. The census projects that people of color will account for over half of the population by 2050 with the largest growth occurring among people who identify as Asian or Hispanic. This shift underscores the importance of refining racial and ethnic categories to capture the complexity of modern identities. Continued adaptation of data collection and reporting methods will be important for reflecting experiences among multiracial people.

Continued efforts to further disaggregate racial and ethnic data may be important for guiding efforts to address health and health disparities. Significant data gaps persist for smaller groups, including AIAN and NHPI people, with very limited information available for subgroups of these populations. Moreover, Asian and Hispanic people are often treated as monolithic in policy discussions, but the diversity in experiences among these groups is vast, encompassing differences in national origin, language, immigration status, and socioeconomic factors that all influence health. Disaggregating data by subgroup can allow for more nuanced understanding of the challenges faced by specific groups and facilitate tailored efforts to address them. For example, the health care experiences of Asian immigrants vary in meaningful ways due to the intersections of race and ethnicity, national origin, income, and other factors that may impact access to health care services. Moreover, data are often key for justifying allocation of resources toward specific communities or groups.

Increased recognition of the intersectional nature of people’s identities and other factors that may affect their health and health care experiences may also have important implications for efforts to address disparities. For example, the combination of race, ethnicity, and gender, highlights disproportionate discrimination for certain groups. Large federal surveys collect demographic and social data, including race and/or ethnicity, gender, educational attainment, and income that can allow researchers to examine how intersectional social and economic factors shape people’s health and experiences. However, in some cases, data are limited to examine experiences by multiple factors. Beyond an individual’s racial or ethnic identity, other factors that are less routinely collected in surveys, such as self-perceived skin color, can also influence their experiences. Socially-assigned race—that is the race that others perceive someone to be—may also be a factor. For example, studies have found that Hispanic or Latino individuals who are socially perceived as White report better health outcomes than those who are perceived as Hispanic or Latino. As efforts to address health disparities continue and evolve, it will be important to consider how these other factors influence people’s experiences and outcomes.

Appendix: Examples of Race and/or Ethnicity Questions Consistent with Revised OMB Standards

Source: Office of Management and Budget, Revisions to OMB’s Statistical Policy Directive No. 15: Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity

Source: Office of Management and Budget, Revisions to OMB’s Statistical Policy Directive No. 15: Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity

1. Between 1790 and 1840 only the heads of free households appeared in early census records, enslaved people were not recorded by name, age, sex, or origin. They only included additional demographic data on White people. ↩︎

Medicaid is the primary payer for long-term services and supports (LTSS) in the United States, and pays for more than two-thirds of the LTSS delivered in home- and community-based settings (HCBS). Most HCBS are optional for states to provide and are offered through “waivers,” which allow states to offer a wide range of benefits and to choose—and limit—the number of people who receive services. The only HCBS that states are required to cover is home health, but states may choose to cover personal care and other services, such as private duty nursing through the Medicaid state plan. States use HCBS waivers to offer expanded personal care benefits or to provide additional services such as adult day care, supported employment, and non-medical transportation. States also use waivers to provide specialized benefits that are specific to the population covered, such as providing supported employment only to people under age 65. KFF estimates that 4.5 million Medicaid enrollees use HCBS, and that the numbers of people using HCBS through the state plan are similar to the numbers using HCBS through waivers. A state’s ability to cap the number of people enrolled in HCBS waivers can result in waiting lists when the number of people seeking services exceeds the number of waiver slots available.

This data note provides new information about waiting lists from KFF’s most recent survey of state Medicaid HCBS programs, including a discussion of why waiting lists are an incomplete measure of unmet need and why they are not necessarily comparable across states or over time, despite KFF’s efforts to obtain as consistent data as possible (see Box 1). Key takeaways include:

• The number of states that maintain waiting lists or interest lists for people who would like to receive HCBS has fluctuated little between 2016 and 2024.
• In most years since 2016, there have been roughly 0.7 million people on waiting lists or interest lists, with a total of over 710,000 in 2024.
• Most people on waiting lists or interest lists have intellectual or developmental disabilities and most live in states that do not screen any people for eligibility prior to adding them to waiting lists.
• Most people on waiting lists or interest lists are eligible for personal care provided through states’ regular Medicaid programs or for services provided through specialized state plan HCBS benefits.
• KFF also recently updated the waiting list indicators on State Health Facts, which show data by state and target population.

Waiting lists provide an indication of people who may need services they are not receiving, but they are an incomplete measure of unmet need because they don’t include people with unmet needs in states that do not cover the applicable services (and therefore, have no waiting list). Waiting lists reflect the populations a state chooses to serve, the services it decides to provide, the resources it commits, and the availability of workers to provide services. In addition, states’ approaches to managing waiting lists differ in how they prioritize and screen for eligibility, making comparisons across states difficult. States are only able to use waiting lists for optional services so the number of people on waiting lists can increase when states offer a new waiver or make new services available within existing waivers; in these cases, the number of people receiving services increases, but so does the number of people on a waiting list. Finally, although people may wait a long time to receive waiver services—40 months on average in 2024—most people are eligible for other types of HCBS while they wait.

Even though HCBS waiting lists are an imperfect measure of unmet need, there are no other alternative measures available. Many HCBS programs were enacted or expanded in response to the Olmstead decision, a court ruling that found the unjustified institutionalization of people with disabilities is illegal discrimination. As 2024 marked the 25^th anniversary of Olmstead, waiting lists are sometimes described as contributing to the risk of unnecessary institutionalization for people with disabilities, and Democrats and Republicans alike proposed legislation in 2024 to address them. Starting in 2027, states will be required to report the number of people on waiting lists as required under a final rule on access to Medicaid services.

How did the number of states with waiting lists change between 2016 and 2024?

Between 2016 and 2024 the number of states with waiting lists has fluctuated between 37 and 41 and is currently at 40 states (Figure 1). While some Affordable Care Act (ACA) opponents have cited waiver waiting lists to argue that expanding Medicaid diverts funds from seniors and people with disabilities, research shows that ACA Medicaid expansion has led to gains in coverage for people with disabilities and chronic illnesses. Waiting lists for HCBS predate the ACA Medicaid expansion, which became effective in most states in 2014, and both expansion and non-expansion states have waiting lists. Waiver enrollment caps have existed since HCBS waiver authority was added to federal Medicaid law in the early 1980s.

There were more changes in waiting lists for specific types of waivers, however. Georgia, Kentucky, and West Virginia reported new lists for seniors/adults with physical disabilities (with a combined 2,683 people) and Alaska, Connecticut, South Dakota, and Washington reported new lists for people with intellectual or developmental disabilities (I/DD, with a combined 13,251 people on the lists). New Jersey newly reported a waiting list of nearly 2,700 people for its Community Care program, which serves people with I/DD and is part of the state’s larger 1115 waiver.

How did the number of people on waiting lists change between 2016 and 2024?

In most years between 2016 and 2024, roughly 0.7 million people have been on waiting lists or interest lists for HCBS (Figure 2). Between 2023 and 2024, total enrollment in waiting lists and interest lists increased by 2.6%. Overall, there was an increase in the number of people on waiting or interest lists in 19 states and a decrease in 14 states.

One factor that contributes to changes over time—especially the notable decline between 2018 and 2020—is that not all states screen for Medicaid eligibility prior to adding people to waiting lists and changes in this policy may result in changes in waiting list volumes. For example, between 2018 and 2020, the total number of people on waiting lists decreased by 155,000 or 19%. However, nearly half of that change came from Ohio’s implementation of a waiting list assessment of waiver eligibility, which reduced the size of the state’s waiting list by nearly 70,000 people. In 2024, most states (32) with waiting lists screen individuals for waiver eligibility among at least one waiver, but even among those states, 4 do not screen for all waivers. The 8 states that do not screen for eligibility among any waivers (Alaska, Illinois, Iowa, Oklahoma, Oregon, South Carolina, Texas, and Washington) account for over half of all people on waiting lists.

In all years since 2016, over half of people on HCBS waiting lists or interest lists lived in states that did not screen people on waiting lists for eligibility. One reason waiting lists provide an incomplete picture of need is that not all people on waiting lists will be eligible for services. Interviews about HCBS waiting lists found that when waiver services are provided on a first-come, first-served basis, people enrolled in waiting lists are in anticipation of future need. That study found that in some states, families would add their children to waiting lists for people with intellectual or developmental disabilities (I/DD) at a young age, assuming that by the time they reached the top of the waiting list, their children would have developed the immediate need for services. Many of those waivers offer comprehensive HCBS packages that include supported employment, supportive housing, or round-the-clock services. Among the eight states that do not screen people for eligibility on any lists, six have only waiting lists, one (Texas) has only interest lists, and one (Washington) uses both. (Illinois does not establish eligibility until selection but does a preliminary evaluation of eligibility prior to placing someone on the list.)

Between 2023 and 2024, several states made changes to their waiting list policies that resulted in notable changes in the number of people on a specific waiting list:

• Illinois implemented a new law requiring school staff to discuss services for children with developmental disabilities in certain situations. The number of people on those two waiting lists increased from 14,444 to 15,905.
• Iowa implemented a new process for assessing people’s need for services to move people with emergent needs higher on the waiting lists for waivers serving seniors/adults with physical disabilities and people with mental health needs. The number of people on the mental health waiting list increased from 802 to 1,128 and on the seniors/adults with physical disabilities waiting list increased from 9,684 to 10,508.
• For its I/DD waiver serving children with Autism, Maryland added a screening call for applicants to ensure that children have an Individualized Family Service Plan or Individualized Education Program, and at least 15 hours each week of special education or related services prior to adding people to the waiting list. The change reduced the number of people on the waiting list from 6,431 to 5,280.

Who is on waiting lists for HCBS?

Most people on waiting lists have intellectual or developmental disabilities (I/DD), particularly in states that do not screen for waiver eligibility before placing someone on a waiting list. People on waiting lists for waivers serving people with I/DD (which include waivers specific to people who have Autism) comprise 89% of waiting lists in states that do not screen for waiver eligibility, compared with 49% in states that do determine waiver eligibility before placing someone on a waiting list (Figure 3). People with I/DD comprise almost three-quarters (73%) of the total waiver waiting list population. Seniors and adults with physical disabilities account for one-quarter (24%), while the remaining share (3%) includes children who are medically fragile or technology dependent, people with traumatic brain or spinal cord injuries, people with mental illness, and people with HIV/AIDS. People who are on HCBS waiting lists are generally not representative of the Medicaid population or the population that uses HCBS. Most people on waiting lists have I/DD, but KFF analysis shows that people with I/DD comprise fewer than half of the people served through 1915(c) waivers (the largest source of Medicaid HCBS spending).

How long do people on HCBS waiting lists wait to access services and do they have access to HCBS while waiting?

In 2024, people on the waiting or interest lists accessed services after an average of 40 months (32 of 40 states responding), down from 45 months in 2021, but up from 36 months in 2023. People with I/DD waited the longest for services, 50 months on average. The average waiting period for other waiver populations ranged from 6 months for waivers targeting individuals with mental illness to 44 months for waivers that serve children. People with I/DD residing in states that do not screen for eligibility wait longer for services than people with I/DD residing in states that do screen for waiver eligibility (70 months versus 43 months, on average).

Most people on waiting or interest lists are eligible to receive other types of HCBS while they wait. Among the 710,000 people on lists for waiver services in 2024, living arrangements are unknown for more than 560,000. Among the people whose living arrangements are known, 98% (147,000) live in the community and 2% (3,300) live in institutional settings. While waiting for waiver services, people living in the community are likely to be eligible for other HCBS through Medicaid state plans. Of the over 4 million people who use HCBS, KFF estimates that roughly half use services provided through the Medicaid state plan, such as personal care to help with bathing or preparing meals, therapies to help people regain or acquire independent living skills, and assistive technology. States may not use waiting lists to restrict the number of people eligible to use such services and over 80% of people on HCBS waiting lists are eligible for personal care or other state plan services. They would not, however, have access to more specialized services such as supported employment or adult day care. People on waiting lists who receive state plan services may also have fewer hours of personal care than they would in a waiver program, or they may not have assistance with some of the activities they need help with such as bathing, dressing, preparing meals, or managing medication.

How are workforce shortages and changes in federal policy affecting HCBS waiting lists?

Although waiting lists may reflect states’ budget constraints, states also use waiting lists to manage shortages of HCBS workers, and with ongoing workforce shortages, it may be difficult to meaningfully reduce waiting lists. Rhode Island and West Virginia newly reported waiting lists because of workforce shortages. In Rhode Island, the waiting list captures people who are eligible for and enrolled in their 1115 waiver but are not receiving home care because of provider shortages. The state noted that provider shortages differ across communities, and that some people waiting for the authorized services might be living in institutions and waiting to transition to community living. West Virginia’s new waiting list for their mental health waiver reflects a lack of workers accepting new patients. Workforce shortages are not unique to Rhode Island and West Virginia, they were reported by all responding states in KFF’s 2023 survey of HCBS programs. It is unlikely that there will be major changes in states’ waiting lists for HCBS without corresponding changes in the availability of HCBS workers.

As states exhaust expanded federal funding for HCBS from the American Rescue Plan Act (ARPA), addressing waiting lists may become more difficult. Section 9817 of the ARPA provided states with an additional 10 percentage points of federal funding for their Medicaid HCBS expenditures between April 1, 2021, and March 31, 2022. States were required to reinvest this additional federal funding into Medicaid HCBS, resulting in an estimated $37 billion in new HCBS funding. As of December 31, 2023, the number one use of the ARPA funds—accounting for more than $26 billion of the planned $37 billion in new funding—was for workforce recruitment and retention. The second largest allocation (an additional $4 billion) was for workforce training. The end of extra workforce-oriented funding could exacerbate workforce challenges, potentially increasing waiting lists, and states planned to spend almost $2 billion directly on reducing or eliminating waiting lists.

The additional ARPA funding will end in most states by March 2025, although 4 states (Missouri, South Dakota, Virginia, and Washington) had already exhausted their funds by August 2024 and 13 states (Alaska, Georgia, Kansas, Maine, Michigan, Montana, North Dakota, New Jersey, New Mexico, Ohio, Pennsylvania, Vermont, and Wisconsin) have received extensions. Three states (California, New Mexico, and Texas), which together accounted for over half of the people on waiting lists in 2024, reported using ARPA funding to reduce or eliminate waiting lists and that continuing those reductions are a top priority. However, it is uncertain how many ARPA initiatives will be sustainable in the long run as the additional federal funds are exhausted.

A new rule on access to Medicaid services will require states to report more information about people waiting for HCBS waiver services starting in July 2027. The new rule requires states to report the number of people who are waiting to enroll in a waiver program, information on whether the people on the list have been screened for eligibility, and the average amount of time people newly enrolled in the waiver over the past 12 months had spent waiting to enroll. Although the regulation does not mention interest lists, referral lists, or registries, the preamble to the rule indicates CMS’ intent for states to report all types of lists.

Recognizing that waiting lists are an imperfect measure of unmet need, the rule also requires states to report two new measures related to access to care for people newly receiving waiver services. For people who began receiving waiver services in the past 12 months, states must also report: (1) the average length of time between approval for homemaker, home health aide, personal care, and habilitation services and the start of services and (2) the percent of authorized hours that were provided. (States may report the latter two measures for a “statistically valid” random sample of recipients.)

Despite the enhanced data states will be required to report, waiting list and waiver information will remain imperfect measures of unmet need. None of the new data will reflect how long it takes for people to receive HCBS provided through the Medicaid state plan, or how comprehensive those services are. The new data also do not capture the number of people whose authorized services are below needed levels because of hourly or dollar caps on the amount of HCBS they can receive. With most people on waiting lists still eligible for state plan services, understanding what services they receive while waiting is also a dimension in understanding how acute the needs are among people on waiting lists. More broadly, the waiting list numbers, and associated requirements, apply to waiver services only, which are optional for states to provide. New data will offer insight into unmet needs among people receiving and waiting for optional services, but there will still be little information about people who need HCBS but are not receiving care.