Vaccination Equity by Design
abstract. This Essay examines how states’ initial COVID-19 vaccine-distribution strategies tended to disadvantage populations of color, including Black, Latinx, and Native American communities. These dynamics resonate with “inverse equity” effects of other public-health innovations. We argue for a federal regulatory framework to reduce inequity-forcing effects during initial vaccine rollout.
Racial disparities in COVID-19 exposure, cases, and outcomes have been a salient feature of the U.S. pandemic. From its earliest days in 2020, COVID-19 transmission skyrocketed in Black, Latinx, and Native American communities. Members of these populations have been two to three times more likely to be hospitalized and more than twice as likely to die from COVID-19 as white, non-Hispanic people.1
As 2020 came to an end, the Food and Drug Administration (FDA) began to approve vaccine candidates for emergency use.2 Plans for vaccination rollout sparked optimism that the pandemic—and the disproportionate burdens it had placed on communities of color—was coming to an end. In the short term, however, optimism about the equity-promoting effects of COVID-19 vaccination was misplaced. In mid-March 2021, surveillance data from forty-four states revealed “a consistent pattern” of lower vaccination rates in communities of color.3 At the national level, data trends showed much of the same: by July 2021, the Centers for Disease Control and Prevention (CDC) reported that of the fifty-seven percent of vaccine recipients for whom race was known, sixty-one percent were white, fifteen percent were Latinx, and just nine percent were Black.4 For Black populations nationwide and in most reporting states,5 rates of vaccination during the initial vaccine-rollout period were lower than the rates of adverse effects from the disease.
These data were no surprise to public-health scholars of the “inverse equity hypothesis.”6 This theory predicts that when health innovations emerge, they are initially adopted by wealthy and connected segments of the population, thereby amplifying rather than reducing inequality.7 COVID-19 vaccination is a case study, characterized by greater uptake in U.S. counties with higher socioeconomic status, lower proportions of racial and ethnic minorities or limited-English-proficiency residents, and greater access to quality housing and transportation.8 These dynamics are part of a familiar pattern, in which socially vulnerable communities—particularly communities of color—already carry a larger proportion of illness and death.9 Vaccines are like other health innovations: they initially widen the gap between the rich and poor. And although these inequality-exacerbating effects tend to abate once there is near-universal access,10 this is no consolation for the communities who experience greater morbidity and mortality in the meantime.
The inverse equity effect of COVID-19 vaccination is predictable, but not inevitable: early vaccine access depends on the design of distribution systems, and these systems can (and should) account for equity concerns early on. Indeed, designers of distribution systems like the CDC and state departments of health had tools to mitigate inverse equity effects from the beginning of the vaccine-rollout period. The progress of vaccination in the United States has been complex, with heterogeneous state distribution programs prioritizing equity, including racial equity, differently.11 At the federal level, regulatory efforts to promote vaccination equity have been incomplete,12 which contributes to the haphazard and often belated efforts to expand vaccine access to the racial groups most at risk.
In this Essay, we identify policy drivers of racial inequity in early access to the COVID-19 vaccine, including health-system infrastructure, physical distribution channels, opt-in and demand-based signups, access to information and distribution platforms, and justified medical mistrust. We then offer a regulatory framework for addressing inequity in vaccine allocation and distribution. Building on literature from behavioral economics, public health, and civil rights, this Essay advocates for the use of federal regulatory tools to counter inverse equity effects from the start of vaccine rollouts. These tools are designed to encourage delivery mechanisms involving trusted local organizations, opt-out systems to supplement or replace demand-based enrollment strategies, and “targeted universalism”13 priorities that broaden access to more disadvantaged individuals within priority populations. Although state governments take the lead in designing U.S. vaccine-distribution strategies, we prioritize federal regulatory tools because they have nationwide applicability and have not been used to their fullest in the COVID-19 vaccine rollout.
We focus on vaccine equity on the basis of race and ethnicity in the United States. In a pandemic context, equitable or “fair” distribution of vaccines could be defined in multiple ways. Most conceptions of equitable distribution suggest that vaccine allocation should track the epidemiological burdens of the COVID-19 disease: vaccines should be distributed in proportions that correspond to each community’s COVID-19-related morbidity and mortality.14 In its weekly Morbidity and Mortality reports, the CDC has defined “vaccine equity” along these lines: “preferential access and administration to those who have been most affected by COVID-19 disease.”15 This differs from what the CDC has called “vaccine equality,” which is characterized by a “similar allocation of vaccine supply proportional to . . . [the] population.”16 Other equity concepts—such as antisubordination (avoiding “practices that enforce the inferior social status of historically oppressed groups”17) and anticlassification (avoiding any practices that “classify on the basis of race”18)—can provide useful shorthand in discussions of vaccine equity, but are less responsive (or unresponsive, in the case of anticlassification) to the immediate allocation of disease burden.19 We will use “equity” to refer to vaccine allocation in proportion to the burden of COVID-19 morbidity and mortality, with a particular focus on racial equity.20
Part I of this Essay introduces the inverse equity hypothesis (IEH) and identifies patterns in early COVID-19 vaccination that appear to show inverse equity effects. Part II examines the factors driving vaccine disparity and shows how key design elements of current vaccine-distribution efforts contributed to or failed to mitigate inequities, relative to alternative designs. Part III draws on civil rights, behavioral economics, and public-health scholarship to propose federal regulatory mechanisms to advance equity.
I. covid-19 vaccination and the inverse equity hypothesis
The origins of the inverse equity hypothesis lie in a 1971 Lancet article, in which British physician Julian Tudor Hart introduced the “inverse care law.”21 The “inverse care law” captures the idea that, in a market-based system, “the availability of good medical care tends to vary inversely with the need for it in the population served.”22 Universal health care through the National Health Service mitigated but did not eliminate this imbalance.23 Hart’s insight has been duplicated across decades, nations, and health-care sectors.24
The IEH is an unhappy corollary to Hart’s inverse care law. Given inequality in the use of existing health-care resources, optimists might hope that new health-care technologies would be allocated in ways that reduce disparities. But this is rarely achieved.25 Studies of innovations as varied as heart surgeries,26 immunizations, HIV treatment, dental-cavity prevention, hospital births, vitamin supplements, and safe water have demonstrated that advances in health-care flow firstto wealthy people, exacerbating rather than reducing disparities.27
According to the IEH, new interventions amplify inequality in a predictable pattern. Early in the diffusion of a new technology, such as a vaccine, wealthy people are early adopters.28 As diffusion expands, demand among wealthier and middle-income people becomes saturated, while the poorest people slowly begin to gain access. Eventually, this delay in access drives persistent disparities to the detriment of the most disadvantaged.29 Vaccines and other innovations thus aggravate inequality when they are first introduced. It is only later, when coverage is widespread and the wealthy have extracted their full benefit, that expanding access to the innovation can narrow health-status gaps between the rich and the poor.30
In the United States, the health gap between the rich and poor tracks longstanding racial disparities in wealth and health care that disadvantage people of color.31 Racial disparities in health care reflect not only structural racism (e.g., socioeconomic exclusion, unequal insurance availability, residential segregation, inaccessibility of health-care facilities, and exposure to incarceration32), but also explicit and implicit racial biases among care providers.33 As applied to the United States, the dynamics of the IEH reflect not only rapid demand for new innovations among wealthier and whiter patients, but also inadequate supplyof new innovations among patients of color due to institutional and interpersonal racism, even when socioeconomic status is held constant.
Pre-exposure prophylaxis medication, which prevents HIV
infection among people who are HIV-negative, provides a case study. Access has
lagged among Black individuals despite the elevated burden of HIV infection
among people of color,34 and implicit racial bias
likely contributes to this effect.35 Many other
studies show lagging access among people of color to health-care technologies,
screenings,36 seasonal flu vaccination,37 pediatric care,38 and organ transplantation.39
Experiences with racism in health care have also prompted justified mistrust among populations of color in the United States,40 particularly given the exploitation of Black people by the medical system since slavery.41 Much of the scholarship on medical mistrust cites to awareness of the Tuskegee Syphilis Study, which withheld penicillin from low-income Black men until as late as 1972 in order to observe the long-term effects of syphilis infection.42 But medical exploitation and neglect of Black communities long predates the events of Tuskegee, and ongoing experiences of discrimination reinforce the view that the health-care system acts against the health interests of Black people.43 Prior work has found that mistrust can cause delays in seeking care and reduced adherence to treatment, and that experiences of health-care discrimination can augment this mistrust.44 In these ways, justified medical mistrust can present an additional access barrier to new health-care technologies among people of color, and it is likely to affect vaccination access in the COVID-19 context.45
The COVID-19 pandemic has generated novel medical
interventions, including tests, treatments, and vaccines. With many of these
innovations, patterns of uptake in the United States and elsewhere have been
consistent with the IEH.46 In some countries, and on a global
scale, this manifests in socioeconomic disparities driven by early access for
the wealthy, followed by lagging access for the most disadvantaged.47 In the United
States, the IEH also reflects racial disparities, demonstrated by early access
for whiter (and wealthier) communities, with lagging access for communities of
color. From the start, racial disparities in infection have been striking, as
have disparities in hospitalization, with people of color more likely to be
exposed, infected, and hospitalized.48 Among
hospitalized patients, several U.S. studies have shown no racial disparities in
mortality, suggesting similar outcomes once patients access hospital
care.49 But these findings focus exclusively “on individuals able to access hospital
care,”50 and racial disparities persist for access to many hospital services.51
Overall, U.S. data have shown that white individuals have accessed COVID-19 vaccination at rates that exceed their share of cases, while Black and Latinx individuals have been vaccinated at lower rates.52 In the first month of vaccine availability, sixty percent of recipients were white, in part due to overall demographics in the prioritized population of health-care personnel.53 But inequitable burdens persist within priority populations. For example, among health-care personnel, a majority of deaths were among people of color, including Black, Latinx, and Asian care workers.54 Some studies have documented higher rates of vaccine hesitancy among Black populations, linked to justifiable medical mistrust.55 However, disparities are present even among people who reported that they were willing to be vaccinated,56 which is an important note given many other differences between people willing and unwilling to receive the vaccine in the United States.57 Racial disparities are one part of overall disparities on the basis of social vulnerability and community disadvantage,58 and findings from December 2020 through April 2021 show that disparities on the basis of social vulnerability increased as vaccine eligibility expanded to include more categories of people.59
These findings are discouraging. But not all health-care diffusion follows the inverse equity path. Governments, health-care systems, and providers can roll out innovations in ways that deliberately reduce inequality from the earliest stages.60 Researchers have sought to identify the features of health-care innovations that lead to inequality-amplifying initial effects. For example, systems using voluntary opt-in approaches can increase inequality, as opposed to opt-out or mandatory policies.61 This is true for COVID-19 vaccination distribution, which requires individuals to actively seek out the vaccine, rather than defaulting to vaccination (i.e., requiring opt-outs). In contrast, interventions that specifically target vulnerable populations—or the use of multiple delivery options—can mitigate inverse equity effects.62 These include, for example, offering financial incentives, reducing financial or access barriers, or specifically tailoring interventions to disadvantaged groups.63
COVID-19 has revealed continuing sources of race-based disadvantage in access to health-care innovations. We now turn to the distribution choices that shaped inverse equity effects in U.S. vaccine distribution.
II. choosing inequitable vaccine distribution
Key elements of the early design and distribution of the COVID-19 vaccines contributed to racial, ethnic, and socioeconomic disparities in initial access. As the IEH predicts, as supply of the vaccine has expanded, disparities have narrowed.64 Yet early differences in vaccination meant that illness and hospitalization rates remained higher in communities of color and low-income communities well after vaccines became available.65 In addition, racial and ethnic disparities in vaccination rates have persisted months after the initial authorization of the vaccine, particularly as measured against COVID-19 deaths.66 These disparities will reverberate beyond the current pandemic because they are missed opportunities to strengthen public-health systems and to increase trust within underserved communities.
After emergency-use authorization of the Pfizer and Moderna vaccines in late 2020, federal regulators needed to tackle the difficult logistical, health, and ethical dimensions of providing access to what was still then a scarce resource. From the outset, federal regulators built in equity considerations when planning for vaccine rollout. The CDC announced its initial guidelines (developed by its Advisory Committee on Immunization Practice (ACIP)) in late 2020 and early 2021. The ACIP was explicit about its goal of reducing disparities given the extensive data on the racial and ethnic distribution of those most affected by COVID-19, as well “inequities in social determinants of health that are linked to COVID-19 risk.”67 Accordingly, an announced goal of the vaccine-allocation framework was to reduce and not exacerbate disparities.68 The guidelines therefore recommended prioritizing access to individuals at high risk because of their age, nature of employment, confinement in long-term care facilities, and medical risk.69 Although this phasing-in system indirectly prioritized many individuals of color by listing high-risk workers, essential workers,70 and those at high medical risk, the guidelines did not explicitly list race, ethnic, or socioeconomic considerations in their allocation plans.71 These guidelines were not binding on states, though most states used key aspects of the CDC framework in designing their own distribution systems.72
Despite this early attention to equity considerations at the
federal level and in many states, the first phase of vaccine distribution
efforts did not avoid the inverse equity trap. Instead, ambiguity in the CDC
guidance and the initial distribution decisions made by many states likely
contributed to race- and ethnicity-based inverse equity effects. Governments
rightly sought to distribute the vaccine rapidly, but they missed opportunities
to mitigate racial, ethnic, and socioeconomic inequities.73 The tendency
by most jurisdictions in the initial distribution stages to rely heavily on
opt-in, demand-based vaccine sign-ups, which depended on information and
technology infrastructures that were already inequitably distributed, was a
critical mistake. These systems interacted detrimentally with longstanding
disparities in access to health care and technology like broadband internet.74 The heavy
reliance on online opt-in systems predictably increased barriers to accessing
the vaccine for communities of color, adding to the obstacles already posed by
justified mistrust in health-care
The remainder of this Part will consider three categories of choices that contributed to inverse equity effects: (A) decisions about the specificity of federal regulatory guidance, (B) state and local decisions to use demand-based and sometimes resource-intensive sign-up systems, and (C) decisions to centralize vaccine distribution in particular geographical locations and institutions.
A. Gaps in the CDC Guidance
First, the CDC prioritization guidelines excluded certain high-risk categories. By relying on broad categories of age, work status, and medical risk for initial prioritization, the guidelines effectively excluded additional ways of measuring social and health vulnerability that might have diminished racial and ethnic impact.76 For instance, the initial guidelines chose not to prioritize communities at high risk or who had suffered high rates of COVID-19 infection and adverse consequences, many of which were communities of color or low-income communities.77
Subsequently, several states adopted prioritization strategies that diverged from the ACIP guidelines. Some states chose to prioritize social vulnerability differently: by high-risk neighborhoods in Rhode Island,78 by multigenerational households in Oregon and Washington,79 and by communities of color (particularly Native American communities) in Montana.80 By the late winter and early spring, as vaccine supply increased, some states targeted particular neighborhoods to reach communities that were most affected by the pandemic, but that had lower vaccination rates relative to the general population.81 This response, while likely helpful, has made visible the initial design choice not to incorporate more direct targeting into early-stage guidelines.
With more specific guidance, the CDC may have prompted more equitable allocation priorities in state vaccination plans. This guidance could have taken several forms. One strategy would have been to keep ACIP’s priority categories (e.g., employment categories82), but to further prioritize the hardest-hit communities within those groups—a strategy that might be thought of as “nested” priorities. Several public-health researchers advocated for this strategy early on,83 based on findings showing racial and ethnic disparities within priority categories. For example, there has been greater mortality among Black health-care workers due to workplace conditions, including the shortage of personal protective equipment (PPE).84 A subcommittee of the National Academies of Science, Engineering, and Medicine (NASEM) recommended designs that embedded priorities even within priority groups—for instance, prioritizing health-care workers of color or workers operating in high-risk communities.85 These designs, however, were not specifically recommended by the CDC.86 A survey of U.S. adults suggests that most people supported approaches tailored to age, mortality risk, and employment, but the survey did not give participants the option to weigh in on race-based prioritization criteria.87 An allocation mechanism that maintains universalist priority groups but also targets vulnerabilities within those groups may be publicly acceptable.
Another more specific strategy—one also recommended by NASEM in response to a request from the National Institutes of Health (NIH) and the CDC to develop a vaccine-allocation strategy for the initial stages of short supply88—would have been to allocate vaccine priority based on measures of social vulnerability. Elsewhere, the CDC has used the Social Vulnerability Index, a measure that has predicted disparities in COVID-19 risk, hospitalization, and death and correlates with race-based disparities.89 NASEM agreed with this measure and recommended prioritizing geographic areas “identified as vulnerable through CDC’s Social Vulnerability Index or another more specific index” for vaccine access.90 Some states approximated measures of social vulnerability in other ways. Rhode Island, for example, incorporated measures of hospitalization, death, and cases to allocate vaccines geographically.91 The state pinpointed these locations, and then made vaccines accessible locally via community clinics, housing, and pharmacies.92
In addition to these strategies, the CDC could have prompted more equitable vaccine distribution by explicitly invoking race and ethnicity alongside other indicators of vulnerability, using a more targeted approach. The faith that employment and age-based priority categories would capture the racially disproportionate impact of COVID-19 ignores the lessons of past universalist approaches that fail to respond to specific mechanisms of subordination or disadvantage.93 Universalist approaches that are nonspecific to race will also miss intersectional categories of greater vulnerability—including race and disability or health risk, race and density, race and place, race and limited English proficiency, and others. When many states moved to age-based categorizations, some commentators94 and state governments realized that these priority groups were not responsive to the racial disparities in disease burden.95
State or federal allocation plans that use race as the sole factor for prioritizing access for individuals could draw constitutional challenges.96 And yet, the use of formally race-neutral criteria, which ignore the distinct role that race and ethnicity has played in disease burden, risks reinforcing disparities in who has access to vaccines, and fails to acknowledge or address vaccine hesitancy linked to justified medical mistrust and histories of racial exploitation by the U.S. medical system. To address this issue, the CDC could explicitly encourage states to attend to racial inequities and barriers to access faced by communities of color. This approach would encourage states to address the ways in which trust, language, insurance, and other factors affect access by people of color. Specific strategies should include partnering with neighborhood and nonprofit groups, community medical- and social-service providers, or faith-based organizations. As Govind Persad has argued, these community-level strategies that consider the specific needs of high-risk communities of color are likely to withstand judicial scrutiny.97 Given the strong correlation between race and COVID-19 disease burden in the United States, issuing CDC guidelines explicitly identifying race as one of multiple indicators of increased vulnerability (and therefore priority) could prompt states to attend to racial equity as an important aspect of distribution mechanisms.
Identifying race would also allow the CDC to make more tailored suggestions to address instances where race gives rise to unique access barriers. For example, a universalist approach does not acknowledge or address vaccine hesitancy linked to justified medical mistrust and histories of racial exploitation by the U.S. medical system.98 This particular barrier will require access strategies that explicitly seek to build trust by partnering with neighborhood and nonprofit groups, community medical and social-service providers, or faith-based organizations. A recommendation by the CDC to include race among multiple indicators of vulnerability would prompt states to adopt frameworks that recognize and address race-specific barriers to access.
B. Demand-Based Distribution
A second key design choice made at the subnational level by many jurisdictions in the initial stages of vaccine distribution was to use opt-in, demand-based approaches to distribute vaccine doses. Individuals seeking a vaccine generally had to register and find appointments themselves; the default was no vaccination, and opting in required up-to-date information and access to opt-in systems. In many jurisdictions, individuals had to schedule appointments using the Internet, rather than via telephone or in-person scheduling or registration, or walk-in appointments.99 This design choice reflected a balance of important considerations, including the need for preregistration to verify compliance with narrow eligibility categories, the need to provide formally neutral and transparent sign-up systems (e.g., a first-come first-serve system within priority groups), and the need to ensure that vaccine registration was not imposed in a way that restricted individual autonomy and choice. But in many jurisdictions, this sign-up system created access barriers for communities of color.
Many sign-up platforms were complicated for all groups to
use, but wealthier and whiter residents had many advantages, including the
necessary equipment, broadband infrastructure, knowledge, and time.100 Communities
of color, low-income communities, and rural communities are less likely to have
Internet and broadband, putting them at a fundamental disadvantage.101 Even widely
publicized voluntarist efforts to increase vaccine access through Twitter and
scheduling bots could not fully address the problem, given that low-income
individuals (and those over the age of sixty-five) are less likely to use
social media platforms.102 As some researchers have
pointed out, the vaccine-rollout process was marked by a sad irony: obtaining a
vaccine often required access to technology, rendering lower-income individuals
who lacked the necessary technological infrastructure even more vulnerable to
Alternative or additional systems such as telephone scheduling, in-person scheduling, and the deliberate integration of community partners or health systems in distribution could have mitigated some of this disadvantage. In the initial stages of rollout, many jurisdictions did not rely extensively on community-based organizations or trusted health systems (namely, sources of care that are less likely to draw medical mistrust, particularly institutions embedded in or operated by local communities, such as neighborhood health centers104) to reach out to potential patients to assist with preregistration and appointment scheduling. Essentially, these strategies would shift from opt-in to opt-out distribution, and from state- or city-run platforms to platforms run by organizations with preexisting credibility and community buy-in.
C. Centralized (Mass) Distribution
A third inequality-amplifying strategy was the overreliance by many states and localities on centralized vaccination sites for initial vaccine distribution. We acknowledge that mass vaccination campaigns can and should be an important part of the pandemic response, but we align with commentators who have noted that mass vaccination should be carried out with attention to “community acceptance, accessibility, and equity,” including communicating with community leaders, using call centers to address language barriers, and making transportation and physical access a priority.105 Locating access to COVID-19 vaccination in large, centralized sites simplifies and speeds up vaccine administration, but also creates distance and travel barriers for at-risk communities, including communities of color.106 Alternative or additional strategies for distribution could have mitigated these impacts, as demonstrated by states and localities that adopted more decentralized or place-based distribution strategies.107 Decentralized distribution strategies that are sensitive to neighborhood characteristics have the dual advantages of reducing logistical access barriers and potentially reducing justified medical mistrust.
Distributing the vaccine through decentralized networks of primary-care providers108—such as community health centers or pharmacies—or in collaboration with trusted community partners could have addressed many place-based and resource-based access barriers, while simultaneously reducing justified medical mistrust.109 Community partners with high social capital, formal and informal resources, and social networks—such as housing providers, faith-based organizations, neighborhood health centers, libraries, pharmacies, and family doctors—can play a crucial role in building trust.110 Peer-reviewed literature supports these distribution mechanisms; for example, a recent systematic review of efforts to reduce vaccine hesitancy has identified logistical and decentralized mechanisms that improve uptake of influenza shots and childhood vaccinations.111 “Dialogue-based interventions”—interventions that work with community and religious leaders, connect with social mobilization efforts, use social or mass media, give local care providers communication tools and training, or push reminders directly to individuals in at-risk areas—showed particular promise, as each of these intervention types was found to increase vaccine uptake.112 Centralized, demand-based distribution systems cannot serve these purposes, and many were discontinued as early vaccine demand was satisfied and the distribution focus shifted to underserved communities.113
The other advantage of decentralized distribution networks, either alongside or in lieu of mass vaccination sites, is that they build capacity. States that initiate and strengthen these efforts, particularly actions involving community partners and care providers (such as community clinics, pharmacies, and health centers), will increase the capacity of these networks to deliver other preventive-care interventions (such as influenza vaccinations and testing for current and future diseases). These benefits can outlast the immediate crisis and contribute to lasting increases in health-care access.
III. equity by design: regulatory tools to avoid inverse equity effects in times of crisis
The prior Part identified alternatives or additions to the strategies that the federal government, states, and localities used to distribute COVID-19 vaccines. The omission of these elements has broader implications for strengthening the regulatory infrastructure and advancing public-health equity. State and local flexibility is essential for a decentralized and tailored approach, given differences in disease burdens, variation in available resources, and the legitimate advantage to be gained from experimentation across jurisdictions.
In this Part, we turn to federal law to identify how federal regulators can encourage (or mandate) vaccine-allocation strategies that would be most effective at reducing inverse equity effects. Federal systems have a comparative advantage in collecting and rapidly analyzing data, publicizing information with credibility, disseminating expertise through guidance, enforcing civil-rights violations, and supporting information networks. In what follows, we suggest a mix of hard and soft federal regulatory approaches to avoid inverse equity effects from the outset.
Throughout this Part, we note that traditional regulatory pacing and deliberation processes are less appropriate in times of widespread infectious disease. Indeed, this is a key reason why emergency statutes temporarily accord greater authority to the executive branch, which enables more nimble responses to emerging threats.114 Some scholars have proposed rapid policy evaluations, which would be useful for informing future emergency responses.115 As a practical matter, however, there is very little time to build in an evaluation ex antein times of crisis. As an ethical matter, there may be insufficient equipoise—the equal likelihood that a new intervention will help or harm its target population compared to the current standard—to use randomized or controlled evaluations. And as an administrative matter, it may be difficult (or inadvisable) to supply different messaging or rules for intervention and control conditions during an emergency.
But other evaluation designs are possible116 and have contributed much to the COVID-19 response. We urge federal, state, and local governments to use these designs to evaluate the equity outcomes of their responses, including inverse equity effects when new preventive or treatment innovations are rolled out.
We now consider five categories of federal regulatory options that could increase equity from the outset of vaccine rollout, even in times of crisis: (1) providing specific guidance regarding equity goals and obligations, (2) providing default equity plans for state adoption, (3) collecting equity outcome data, (4) publicly disseminating and ranking state equity outcomes, and (5) facilitating information sharing among states.
The framework we suggest below is well within the power of federal authorities. Undoubtedly, the pandemic raised fundamental questions about the role of federal, state, local governments in regulating public health. The early days of the pandemic were marked by sharp contestations over distribution of PPE, stay-at-home orders, and mask mandates that implicated federalism and localism. And some have argued that the pandemic demonstrates the need for a more centralized federal authority over public-health emergencies.117 While states have police power over public health,118 the specific recommendations below flow from the CDC’s and the Department of Health and Human Services’ (HHS) current statutory and regulatory power over vaccine distribution, interstate disease transmission, data collection, and the conditioning of federal funds to prohibit discrimination.119
A. Equity Directives
Federal systems designers should have attended to race, ethnicity, and social vulnerability more explicitly in the initial guidance that they issued to states and localities. There are at least two points in the regulatory framework where federal agencies might have advanced racial and ethnic equity in the distribution of COVID-19 vaccines. First, as suggested in Part II, the CDC’s allocation guidance could have recommended that jurisdictions attend to racial and ethnic disparities within priority groups, and they could have encouraged jurisdictions to prioritize counties or neighborhoods that were socially vulnerable and most affected by the pandemic.
Second, HHS and the Department of
Justice (DOJ) should have issued clear
directives under Title VI of the 1964 Civil Rights Act and other relevant
civil-rights statutes to require states to
adopt practices that would have promoted equitable vaccine
access.120Title VI extends to federally funded programs,121 and the applicability of Title VI and other civil-rights statutes to funding in health care was recently reinforced by the Affordable Care Act.122 Federal assistance for the COVID-19 response, including the funds made available to support testing, vaccination, and treatment, likely would have qualified as “health program[s] or activit[ies],” enabling the DOJ or the HHS Office for Civil Rights to enforce Title VI in state COVID-19 vaccination rollouts.123 Title VI also imposes duties of inclusion that go beyond nondiscrimination to require the elimination of unjustified barriers.124 And in other contexts, federal agencies have used their power under Title VI and federal civil-rights statutes to ensure that federal spending operates to diminish inequality and promote fair access to federally subsidized programs.125 Indeed, HHS issued affirmative inclusionary guidance on the location and accessibility of COVID-19-testing sites on the basis of race, ethnicity, and language.126 The agency also issued disability guidance requiring accommodation and standards of accessibility in the physical structure, location, and aspects of vaccination programs.127
Similar affirmative guidance from HHS and the DOJ might have directed states to avoid COVID-19 vaccine-distribution mechanisms that exacerbate racial and ethnic inequities, while instructing them on how to improve access. This could include advising grant recipients to locate vaccination sites in neighborhoods of color most affected by the pandemic or in decentralized accessible sites such as pharmacies or large housing providers, to provide transportation or ensure accessibility via public transit, to prioritize residents from particular census tracts for sign-ups, to allow walk-in or telephone scheduling as an alternative to internet-based registration, to ensure that registration and outreach materials are available in multiple languages, and to partner with community groups for outreach to particularly hard-to-reach populations.128
In addition to guiding states and localities in their vaccination plans, federal agencies can implement these strategies directly in vaccination centers that are federally funded and operated. COVID-19 federal vaccination efforts have included a retail pharmacy program and direct provision of vaccine doses and funding to community health centers in key areas.129 Numbers released by the White House suggest that as of March 2021, federal vaccination sites and federally funded community health centers provided sixty to sixty-five percent of their allocated COVID-19 vaccines to people of color.130 These federal distribution programs have been funded by the CDC and by the Health Resources and Services Administration (HRSA), another section of HHS.131 Mass vaccination programs implemented by the Federal Emergency Management Authority (FEMA) have been part of these efforts, and commentators have suggested ways to make these programs more equity focused by adjusting allocation processes to prioritize underserved groups.132
B. Default Equity Plans
One means of inducing states to attend to racial equity is to require them to file a vaccine-equity plan with an agency, like the CDC, that employs scientific experts with the skills to evaluate such plans and make recommendations. As part of the CDC’s response to racial disparities in COVID-19 burdens, the agency appointed a Chief Health Equity Officer and began specifically collecting and reporting data on COVID-19 outcomes by race and ethnicity.133 The agency did not, however, require grant recipients to commit to specific equity-enhancing strategies. To enforce such a requirement, Congress could draw on its Spending Clause authority to attach conditions to financial assistance, such as federal aid to vaccination programs.134 When Congress delegates authority to HHS or the CDC to administer state funding, the agency might also possess implied (or explicit) statutory authority to set application requirements. The congressional acts addressing COVID-19135 have allocated more than $67 billion to the CDC for its COVID-19 response,136 including more than $20 billion to support the distribution and tracking of coronavirus-vaccine distribution.137 The CDC has now distributed over $55 billion to states, tribal governments, local governments, and territories in grants and assistance, including more than $7 billion earmarked for vaccination efforts.138 This includes a series of grants specifically intended to promote equity, for which “75% of the total funding must focus on . . . vaccine access, acceptance, and uptake among racial and ethnic minority communities; and 60% must go to support local health departments, community-based organizations, and community health centers.”139 If the CDC had required an equity plan as part of the application or reporting for funding, states would have incentives ex ante to build racial equity concerns into their vaccine rollout plans.
We propose, however, that the CDC go further than simply asking for an upfront equity plan, and instead propose a default plan that states can choose or modify. Although some state-by-state tailoring will be important, the CDC could offer states a default equity strategy as a starting point. States could then select the parts of the strategy that they wished to pursue or adopt the default plan wholesale. The creation of a default plan could save states time as they seek federal funds, and ensure that states at least consider implementing strategies that promoted equity effectively in the past. This strategy would not entail withdrawing funds if states departed from their plans. But identifying and publishing states’ equity strategies would ensure that states at least consider equity issues, and it would harness reputational incentives for states to follow through with their selected strategies.140
C. Data Collection
Data are key to understanding disparities and advancing equity. A core function of the CDC is to conduct surveillance of infectious diseases at the federal level.141 The process of surveillance, however, relies on voluntary participation by states. All states require reporting of reportable disease cases, with reportable diseases identified by the state itself.142 Further, the CDC provides states and localities guidance and resources to aid them in their reporting.143 Where the CDC has also designated a particular disease or infection as notifiable,144 states then provide these reports voluntarily to the CDC, which collects information via the National Notifiable Disease Surveillance System.145 In times of emergency, the Public Health Services Act allows the CDC to waive the need for additional approvals to request state information.146 Therefore, although state data reporting to the CDC is still voluntary in emergency circumstances, the agency has more flexibility in determining what information to request.147
Congressional responses to COVID-19 have allocated additional funding to the CDC to support the agency’s Data Modernization Initiative, an effort to improve disease surveillance in collaboration with state, local, and territorial governments.148 The Coronavirus Aid, Relief, and Economic Security (CARES) Act allocated $500 million to the CDC to improve data infrastructure,149 with particular efforts to support automated data reporting and electronic transmission of data.150 If these systems succeed in streamlining the administrative burdens that may deter states from reporting notifiable diseases by race, this will be a critical public-health investment for monitoring equity in future pandemics.
The CDC has also made efforts to use its grant authority to secure improved reporting of vaccination data by race and ethnicity. In its 2020 Interim Playbook for Jurisdiction Operations for the COVID-19 Vaccination Program, the CDC directs funding recipients to “address all requirements outlined in the playbook and clearly describe their responsibility for ensuring activities are implemented.”151 In order to receive or administer vaccinations, provider facilities supported by CDC funds must enroll in the federal COVID-19 Vaccination Program, which requires an agreement to “record and report required information” for every vaccine dose administered.152 This required information, in turn, includes a record of the recipient’s race and ethnicity.153 These types of reporting requirements are a useful avenue for collecting race-specific data. Although this current arrangement extends only to COVID-19 vaccine distribution in CDC-funded programs, regulators could use this strategy more broadly to routinely mandate demographically disaggregated data on infections and deaths.
The CDC should routinely request information from states that disaggregates both process and outcome data by race and ethnicity.154 When President Biden took office, his administration placed an increased focus on health equity with regard to COVID-19. This included establishing a COVID-19 health equity taskforce, making recommendations to increase data collection that reflected race and ethnicity, and improving responses tailored to the COVID-19 needs and challenges of particular populations.155 An additional and clarifying step would be for Congress to authorize the CDC to mandate that states report on diseases and vaccination explicitly by race or neighborhood, even if the agency could only exercise that power during an active public-health crisis.
D. Dissemination and Ranking Equity Outcomes
In addition to collecting data on equity-related outcomes, we would also urge the CDC to report these data publicly on a state-by-state level, allowing for the comparison and ranking of states. Multiple metrics could be used to accommodate different equity definitions, such as vaccination rates as a proportion of overall disease burden, as a proportion of population share, or as a proportion of mortality. Disclosures employing these metrics would increase public focus and reputational motivations for states to improve the equity outcomes of vaccine rollout, and it would give community-based advocates the information needed to encourage improvements. At the start of the rollout, a number of states were not sharing vaccination data disaggregated by race.156 Efforts to track equity outcomes relied on private actors such as the Kaiser Family Foundation157 or news outlets.158 As of June 2021, however, most states have included race and ethnicity information on their public COVID-19 dashboards, though states vary in the categories they collect and highlight.159 Routine CDC reporting could encourage states to post their own data by race and ethnicity categories, which would increase transparency. The CDC reporting of state data would be highly visible, and it would boost the legitimacy, utility, and accuracy of state information. The CDC does, in fact, report these data for the United States in aggregate (rather than by state or county, at the time of this writing), but it is limited by variability and incompleteness in state reporting.160
This recommendation would be particularly powerful in conjunction with state-provided equity plans, as described above. Alongside each state’s equity outcomes, the CDC could also post the state’s equity plan. If many states adopt similar strategies or select options from the default plan, the agency could check off each state’s choices from a list of common strategies. This would facilitate comparisons between states, and it would give advocates more insight into the types of equity-promoting strategies that states themselves have planned. It may also help track which strategies are more effective in promoting equity across states, which can inform evaluations and ongoing adjustments.
Transparency in reporting is necessary, but insufficient. Experimental studies suggest that adults in the United States are accustomed to racial disparities, such that publicizing even gross disparities does not draw sustained engagement or public outrage. Worse, it can invite people to rationalize disparities.161 Still, even though transparency cannot remedy racial inequity by itself, it is an important complement to all other equity-promoting strategies.
E. Facilitating Information Sharing
Just as the CDC can be a credible hub for vaccine-equity outcome data, it can also provide a platform for states and localities to share strategies that may be equity promoting (or that turn out to be equity inhibiting). The CDC has long run a program identifying “effective behavioral interventions (EBIs)”—programs that have demonstrated effectiveness for preventing HIV.162 The agency actively identifies programs for different population groups, designates them as EBIs, and posts information that helps public-health actors identify and adopt those programs in practice.163 The agency also provides technical assistance to new program implementers.164
Although the full measure of this approach may not be feasible in pandemic times, the CDC can provide web space for states and localities to offer equity-promoting strategy ideas, create a forum for the public to suggest ways to increase equitable distribution, or facilitate information sharing among states (such as through webinars or discussion groups).165 The CDC could publicize best or promising vaccine-equity practices for localities, states, and the public, similar to the efforts the agency has made to identify evidence-based HIV prevention programs. Encouraging the diffusion of best practices could reduce duplicative efforts, increase problem solving, and ultimately promote equitable vaccine distribution.
Improving racial equity in vaccine access is a legal, ethical, and practical obligation, particularly in a crisis. In this Essay, we have identified the tendency for initial vaccine rollout to be inequality-forcing, rather than equity-building, drawing on a large body of public-health research documenting the inverse equity hypothesis. COVID-19 vaccination has fallen into this familiar pattern of early access by the wealthy and white, followed by lagging access by people of color and the most disadvantaged. We have analyzed ways in which federal agency guidance and states’ own choices have contributed to exacerbating vaccine inequity, which in turn translates into disparities in COVID-19 disease burden. And we have demonstrated that these choices were not inevitable at the outset and are not unchangeable now. Federal regulatory tools are available to increase equitable vaccine distribution, many of which leverage existing authorities that have gone untapped in the COVID-19 response.
Our focus in this Essay has been racial and ethnic equity, which is critical given the distribution of COVID-19’s burdens. But these strategies can also promote equity along other axes, such as disability, which is already the subject of specific CDC guidance.166 For any underserved group, it would be helpful to conceptualize equity as measured by vaccine access in relation to the burden of disease, rather than alternatives such as equal lottery chances of vaccination or access in relation to population proportion.
For many of our proposals, an enduring benefit of promoting equitable COVID-19 vaccine distribution is the development of systems and capacity to deliver health care more equitably in nonpandemic times. Building trust through community partnerships, local health infrastructure, transparent reporting, and mobile or decentralized vaccine delivery can provide lasting benefits for access to other types of health care.
As COVID-19 and other infectious diseases persist, we will continue to learn what works (and what doesn’t) for rolling out new interventions equitably. By using regulatory tools and systems designs that build in equity from the start, we have a much better chance of mitigating inverse equity effects now and in the next crisis.
Professor of Law, Cornell Law School (Underhill); Jerome B. Sherman Professor of Law, Columbia Law School (Johnson). We are grateful to I. Glenn Cohen, Govind Persad, Bill Sage, and participants in the Cornell Faculty Workshop (including Sara Bronin, Kevin Clermont, Sherry Colb, Deborah Dinner, Michael Dorf, Maggie Gardner, Valerie Hans, George Hay, Jaclyn Kelley-Widmer, Odette Lienau, Jeff Rachlinski, Aziz Rana, Jed Stiglitz, Nelson Tebbe, and Ezra Young) for helpful comments on this work. We note that COVID-19 vaccine-distribution strategies and outcomes will change over time; we focus here on the early stages of rollout, where inverse equity effects should be most visible. Professor Johnson’s research for this Essay was funded in part by the David J. Greenwald Faculty Research Fund.