Profound racial and ethnic disparities in health and well-being have long been the norm in the United States.
Black and American Indian/Alaska Native (AIAN) people live fewer years, on average, than white people. 1 They are also more likely to die from treatable conditions; more likely to die during or after pregnancy and to suffer serious pregnancy-related complications; and more likely to lose children in infancy. 2 Black and AIAN people are also at higher risk for many chronic health conditions, from diabetes to hypertension. 3 The COVID-19 pandemic has only made things worse, with average life expectancies for Black, Latinx/Hispanic, and, in all likelihood, AIAN people falling more sharply compared to white people. 4
People’s health also varies markedly across and within states, as does access to health services and overall quality of care. 5 Large racial and ethnic health inequities, driven by factors both inside and outside the health care delivery system, are common. In many communities of color, poverty rates are higher than average, residents tend to work in lower-paying industries, and residents are more likely to live in higher-risk environments — all contributors to COVID-19’s disproportionate impact. 6
Issues around cost, affordability, and access to care also contribute to inequities. Black, Latinx/Hispanic, and AIAN populations are less likely to have health insurance, more likely to face cost-related barriers to getting care, and more likely to incur medical debt. 7 It is also less common for individuals from these groups to have a usual source of care or to regularly receive preventive services like vaccinations. 8 In addition, many people of color contend with interpersonal racism and discrimination when dealing with clinicians and more often receive lower-value or suboptimal care. 9
Decades of policy choices made by federal, state, and local leaders have led to structural economic suppression, unequal educational access, and residential segregation, all of which have contributed in their own ways to worse health outcomes for many people of color. 10 The failure to ensure all Americans have reliable health coverage has paved the way to inequitable access to health care. Dramatic disparities in the quality of health care, meanwhile, are tolerated. And while the effects of structural racism persist in all states, 11 policy leaders in some states are reluctant to take actions that could mitigate health inequities, like expanding eligibility for Medicaid as provided for under federal law. 12
The Commonwealth Fund’s Scorecard on State Health System Performance has long tracked the functioning of each state’s health care system, with the goal of motivating actions to improve their residents’ health and health care. But assessing how well a state’s health system performs on average can mask profound underlying inequities.
In this report, we evaluate health equity across race and ethnicity, both within and between states, to illuminate how state health systems perform for Black, white, Latinx/Hispanic, AIAN, and Asian American, Native Hawaiian, and Pacific Islander (AANHPI) populations. Our hope is that policymakers and health system leaders will use this tool to investigate the impact of past policies on health across racial and ethnic groups, and that they will begin to take steps to ensure an equitable, antiracist health care system for the future.
Our measurement strategy was designed to produce a state health system performance score for each of five racial and ethnic groups in every state where direct comparisons are possible among those groups as well as among groups in other states.
We started by collecting data for 24 indicators of health system performance, stratified by state and by race and ethnicity. Indicators were grouped into three performance domains: 1) health outcomes, 2) health care access, and 3) quality and use of health care services.
Scoring method. For each of the 24 indicators, we calculate a standardized z-score for each state/population group with sufficient data. As an example, for adult uninsured rates, we calculate standardized scores using point estimates for 191 pairs of state racial and ethnic groups (51 white, 48 Latinx/Hispanic, 39 Black, 37 AANHPI, 16 AIAN) with sufficient data.
Within each performance domain, we combined indicator values to create a summary score. We then combined the domain summary scores to create a composite state health system performance score for each racial and ethnic group within a state — Black (non-Latinx/Hispanic), white (non-Latinx/ Hispanic), AIAN (non-Latinx/Hispanic), AANHPI (non-Latinx/Hispanic), and Latinx/Hispanic (any race). The ability to generate these scores is dependent on having a sufficient population sample size for analysis.
Based on the overall composite scores, each racial/ethnic group within each state received a percentile score providing both national and state-level context on the performance of a state health system for that population.
The percentile scoring, from 1 (worst) to 100 (best), reflects the observed distribution of health system performance for all groups measured in this report and enables comparisons within and across states. For example, a state health system score of 50 for Latinx/Hispanic individuals in California indicates that the health system is performing better for those residents than Latinx/ Hispanic people in Florida, who have a score of 38, but worse than white residents in California, who have a score of 89. It is important to note that because scores are set relative to one another rather than to a predefined benchmark, there is still room for improvement in health system performance at or near the 100th percentile.
Use of racial/ethnic data categories. The five racial and ethnic data categories we include in this report often group together populations with different experiences, cultures, immigration barriers, and other socioeconomic factors. This includes a wide range of culturally distinct Latinx/Hispanic communities and Asian American communities. Such groupings are imperfect, as they mask significant and important differences. For example, past research has shown variability in health insurance coverage rates among Asian American subpopulations and between Asian Americans and Native Hawaiians or Pacific Islanders. 13
Use of these categories is necessary to obtain sufficient sample sizes for analysis. But states and localities should interpret the findings within the context of their own communities, using them as a starting point to help guide more targeted research and policy solutions.
Refer to the appendix for complete study methods, list of indicators, and health system performance scores for each state’s and racial and ethnic populations.
Both across states and within states, health care system performance varies widely by race and ethnicity, as shown in Exhibit 1. Mirroring the nation as a whole, substantial health and health care disparities exist between white and Black, Latinx/Hispanic, and AIAN communities in nearly all states.
Even in states that achieve high performance overall, racial and ethnic disparities can be dramatic. For example, Minnesota, which ranked third in the Commonwealth Fund’s most recent State Scorecard on Health System Performance, has some of the largest disparities between white and Black, Latinx/Hispanic, AANHPI, and AIAN communities. 14 Some states, like Mississippi, demonstrate relatively poor performance for all groups.
In the small number of U.S. states where AIAN communities represent a sizeable portion of the nonwhite population — such as South Dakota and Alaska — wide performance gaps are also apparent. While the health system in many states tends to perform better for AANHPI populations, performance is lower in New York and Texas, home to two of the country’s largest AANHPI populations.
The overall health system score for each group within a state represents the aggregate performance across three dimensions: Health Outcomes, Health Care Access, and Quality and Use of Health Care Services. Below we describe findings for each of these domains.
Health outcomes, as measured primarily by mortality rates and the prevalence of health-related problems, differ significantly by race and ethnicity. In most states, Black and AIAN populations tend to fare worse than white, Latinx/ Hispanic, and AANHPI populations. While enduring lower life expectancies for Black and AIAN individuals in the U.S. can be attributed in large part to generations of structural racism, oppression, and other factors beyond health care delivery, the health care system nevertheless has a crucial and often unfulfilled role in mitigating disparities. 15
We can get a glimpse of the care delivery system’s role in these unequal outcomes by looking at the frequency of deaths before age 75 from preventable and treatable conditions — a measure known as mortality amenable to health care that is highly correlated with life expectancy. 16 In nearly every state, Black people are more likely than white people to die early from preventable causes (Exhibit 2). Latinx/Hispanic individuals, however, generally have lower preventable mortality rates, despite their comparatively poor access to health care. These lower rates could be related to immigration factors, to a younger average age, or to lower rates of risky health behaviors like smoking. 17 Still, recent research shows increasing mortality and prevalence of chronic conditions for Latinx/ Hispanic populations. 18 There are also differences in outcomes between different Latinx communities. 19
We also see distinct regional patterns. For example, preventable mortality rates are higher for both Black and white residents of many southeastern states compared to other parts of the country, while rates among AIAN people tend to be higher in the upper Midwest and northern Plains states. Among Latinx/Hispanic people, premature mortality rates are higher — and align more closely with rates among white people — in several southwestern and mountain states, including Arizona, Colorado, New Mexico, Oklahoma, Texas, and Wyoming.
Diabetes is an example of a disease that can often be effectively managed — for example, with consistent blood glucose monitoring and proven medications — but is nonetheless associated with profound racial and ethnic disparities in outcomes. Black and AIAN individuals are much more likely to die from diabetes-related complications (Exhibit 3) than people of other races and ethnicities. Health systems striving for equity should bolster disease management resources among these communities to achieve better outcomes.
We also see sizeable disparities when looking at mortality rates for other treatable conditions. Breast cancer, for example, is often considered treatable when detected early but is more likely to be diagnosed at later stages in Black women, who have much higher age-adjusted death rates for the disease across most states compared to other women (Exhibit 4). 20 Across all education levels, infant and maternal mortality rates are higher for Black and AIAN residents than for others. 21
States can perpetuate disparities by not removing barriers to people receiving preventive services, getting effective treatment for chronic conditions like diabetes and high blood pressure, and receiving coordinated care. These barriers range from poor insurance coverage, lack of a usual source of care, and unaffordable medications, to clinicians who prescribe less-effective services or fail to provide timely care for a chronic disease. 22 Sometimes differential outcomes also can reflect unequal access to higher-performing providers, but disparities in care occur even within the same provider facilities. 23
Large disparities in access to care between white and most nonwhite populations are apparent across states. Latinx/Hispanic people typically face the highest barriers to care, although, as noted above, they also tend to have better health outcomes than many other groups (despite variations by geographic region).
A key contributor to these access inequities is lack of comprehensive insurance coverage, or any coverage at all. Insurance alone cannot guarantee access, but it is necessary for getting needed health care without incurring substantial or even catastrophic financial risk.
Americans get their health coverage either from commercial insurance plans offered by employers or sold in the individual market, or from public insurance programs like Medicaid, Medicare, and the Children’s Health Insurance Program. Prior to the Affordable Care Act (ACA)’s major coverage expansions in 2014, limited access to employer health benefits, more restricted eligibility for Medicaid, and often unaffordable individual market plans created significant inequities in coverage among adults.
After the health law’s coverage expansions, adult uninsured rates fell across all racial and ethnic groups. Still, in nearly all states, uninsured rates continue to be higher for Black, Latinx/Hispanic, and AIAN people than they are for white people (Exhibit 5).
Some Latinx/Hispanic and AANHPI populations continue to face immigration-related barriers to getting enrolled in coverage through Medicaid or the ACA marketplaces. While American Indians and Alaska Natives can obtain certain health care services through the Indian Health Service (IHS), lack of insurance coverage can hinder access to needed care outside of persistently underfunded IHS facilities. 24
The ACA created a federal standard for comprehensive insurance and provides for subsidized coverage through marketplace plans and Medicaid. But 12 states have yet to take advantage of the law’s expansion of Medicaid eligibility, which has significantly improved equity in coverage and access and has helped health care facilities in underserved communities (including IHS providers) become more financially stable. 25 Further, Black and Latinx/Hispanic communities are disproportionally represented in states that have not expanded Medicaid: 43 percent of Black and 36 percent of Latinx people live in the 12 nonexpansion states.
When people are uninsured, experience gaps in coverage, or are in private plans that do not provide comprehensive coverage, they often avoid getting care when they need it or pay high out-of-pocket costs when they do seek care. 26 This is particularly burdensome for individuals with lower income and little wealth — disproportionately people of color. 27 Because of these costs, Black, Latinx/Hispanic, and AIAN people are more likely to avoid getting care when they need it, more often have higher out-of-pocket costs, and are more prone to incur medical debt at all income levels. 28
The proportion of white people reporting cost as a barrier to receiving needed care ranges from 6 percent in the District of Columbia and Hawaii to 14 percent in Georgia, Oklahoma, Alabama, and Mississippi. But among Latinx/ Hispanic people, state rates vary between 10 percent in Hawaii to a high of 30 percent in Tennessee (Exhibit 6).
Many people of color in the U.S. are also less likely to have a usual source of care, an important point of contact with the health system that can help people get treatment when they need it. Lack of a regular care provider often goes hand in hand with high uninsured rates and high patient cost sharing. But it also reflects low Medicaid payment rates that limit the network of participating providers and hospitals, a lower concentration of providers and health facilities in neighborhoods where people of color reside, and language and cultural communication barriers. 29 For AIAN communities in rural areas, who are among the least likely to have a usual source care, geographic barriers can also be a key factor. 30
Racial and ethnic disparities in the quality of care and the use of services have also been extensively documented. Across and within most states, white populations overall receive better care than Black, Latinx/Hispanic, American Indian/Alaska Native (AIAN), and, often, Asian American, Pacific Islander, and Native Hawaiian (AANHPI) individuals.
Primary care clinicians play an especially critical role in providing people with high-value services, including preventive care like cancer screenings and vaccines, as well as chronic disease management. When there are barriers to obtaining primary care, people are more likely to get care in more intense and costly care settings, particularly an emergency department (ED).
On two measures of primary care effectiveness, Black Medicare beneficiaries are more likely than white beneficiaries to be hospitalized for acute exacerbations of treatable and manageable chronic illnesses and to seek and receive care in an ED for conditions that are nonurgent or treatable by a primary care provider (Exhibit 7). For both Black and white Medicare beneficiaries, more primary care spending is associated with less use of the ED for treatable conditions and fewer hospital admissions. 31
Primary care settings are also where the majority of vaccinations in the U.S. have taken place, and they play an important role in COVID-19 vaccination efforts. On average, Black and Latinx people are less likely than white people to have received recommended vaccines. In 2019, Black and Latinx children were less likely than white children to have received all of seven key vaccines by age 35 months, but differences were relatively small. Conversely, less than half of all adults received an annual flu shot in 2019–20, and racial/ethnic inequities are apparent (Exhibit 8). Strong federal policy can help close these gaps. For example, the Vaccines for Children program run by the Centers for Disease Control and Prevention (CDC) promotes early childhood vaccination and makes vaccines available at no cost to a partner network of state and local health departments. This, along with state polices regulating vaccination, have proven successful for raising vaccination levels for all children. 32
Expanded access to primary care improves health outcomes. And given the relatively lower use of primary care by Black, Latinx/Hispanic, and AIAN people, these groups in particular are likely to see a greater health impact from improved access and quality.
Racial and ethnic disparities in health outcomes and health care are pervasive both across and within states. Transformative change will depend on policy and practice changes to make access to care more equitable and to ensure equal treatment in the delivery of care.
While health systems alone cannot address all the structural inequities that contribute to differential health outcomes, there are a number of policy options for addressing unequal access to care and unequal treatment within health care facilities.
We group these federal and state policy priorities into four areas:
Ensuring universal, affordable, and equitable health coverage. Nearly 30 million people in the United States are still uninsured, and they are disproportionately people of color. Even those who have some coverage face rising levels of financial risk. Policy options include:
Strengthening primary care and improving the delivery of services. Communities that are predominantly Black and Latinx/Hispanic tend to have fewer primary care providers and lower-quality health care facilities than communities that are mostly white. 40 Federal and state policymakers could start to reverse these inequities by raising payment for primary care providers and transitioning primary care reimbursement to value-based payment that enables investment in health promotion, disease prevention, and chronic disease management. 41 For example, North Carolina now has a prospective Medicaid payment model that emphasizes primary care–based population health management, while Oregon and Washington are linking Medicaid payments to performance on equity measures. 42
There are also opportunities to change how care is delivered and who delivers it:
Reducing inequitable administrative burdens affecting patients and providers. Americans seeking health care face far higher administrative hurdles than residents of other high-income nations. 45 Recent research points to the negative impact these barriers have on access to care for lower-income individuals, including many people of color. 46 Autoenrollment is one reform that could reduce the application burden associated with state Medicaid programs; it could help people get, and stay enrolled in, public coverage. 47 If poorly designed, the quality reporting, care management, utilization review, and prior authorization programs instituted by public and private insurers can create unnecessary red tape and even financial penalties for underresourced providers. Administrators could audit oversight and accountability programs for their disproportionate impact on providers serving communities of color.
Investing in social services. The U.S. spends less on economic and social supports for children and working- age adults than most other high-income countries, and the lack of adequate investment in this area likely contributes significantly to racial and ethnic inequities in health outcomes. 48 Federal and state policymakers could expand economic support for lower-income families by implementing unemployment compensation and Earned Income Tax Credit and child tax credit programs, as well as childcare, food security, and targeted wealth-building programs. 49 Additional investments in affordable housing, public transportation, early childhood development, and affordable higher education also could help reduce racial and ethnic health inequities. 50
Racial and ethnic equity in health care should be a top priority of federal and state policymakers. A good start would be to identify policies and proposed legislation that impede progress toward health equity.
Given that structural racism has played a significant role in shaping those policies that have spawned widespread health inequities, leaders at the federal, state, and local levels should reexamine existing laws and regulations for their impact on people of color’s access to quality care. And new reforms to ensure good insurance coverage and timely access to primary and specialty care need to target communities across the United States that have long been ignored.
Equally important is the development and use of equity- focused measures to monitor the progress of efforts intended to advance health equity and to engender accountability for achieving desired outcomes. And systems are needed to track whether states, health systems, and health plans are reducing racial disparities in clinical outcomes, coverage, access to clinicians, and a host of other health-related gaps.
Too often in the U.S., race and ethnicity are correlated with access to health care, quality of care, health outcomes, and overall well-being. This is a legacy of structural, institutional, and individual racism that predated the country’s founding and that has persisted to the present day, in large part through federal and state policy. By pursuing new policies that center racial and ethnic equity, expand access to high-quality, affordable care, and bolster the primary care workforce, we as a nation can ensure that the health care system fulfills its mission to serve all Americans.
We owe our sincere appreciation to the four member advisory panel who provided crucial feedback and review throughout development of the methods used in this report — the group included Cara James, Ph.D. (Grantmakers In Health); Zinzi Bailey, Sc.D., M.S.P.H. (University of Miami Miller School of Medicine); Dolores Acevedo-Garcia, Ph.D., M.P.A.-U.R.P. (Brandeis University); and Marc Elliott Ph.D., M.A. (RAND Corporation).
We would also like to thank the researchers who developed indicators and conducted data analyses for this scorecard. These include: Michael E. Chernew, Ph.D., and Andrew Hicks, Department of Health Care Policy, Harvard Medical School; Sherry Glied, Ph.D., and Mikaela Springsteen, New York University Robert F. Wagner Graduate School of Public Service; and Angelina Lee and Kevin Neipp, Westat.
We would like to thank the following Commonwealth Fund staff: David Blumenthal, M.D., Melinda Abrams, and Rachel Nuzum for providing constructive guidance throughout; and the Fund’s communications and support teams, including Barry Scholl, Chris Hollander, Deborah Lorber, Bethanne Fox, Josh Tallman, Jen Wilson, Paul Frame, Naomi Leibowitz, Samantha Chase, Relebohile Masitha, Arnav Shah, Aimee Cicchiello, Christina Ramsay, Alexandra Bryan, Sara Federman, Munira Gunja, and Celli Horstman for their guidance, editorial and production support, and public dissemination efforts.
Finally, the authors wish to acknowledge Maya Brod of Burness Communications for her assistance with media outreach, and Westat for its support of the research unit, which enabled the analysis and development of the scorecard report.